Study on the Parameters of Ice Clouds Based on 1.5 µm Micropulse Polarization Lidar

Round 1

Reviewer 1 Report (Previous Reviewer 1)

The authors have extensively corrected all of the suggestions and questions that I had with the original manuscript which primarily were associated with parameter and technical specifications of the lidar system and other instrumentations.   The authors have also answered other questions from reviewers concerning aerosol parameters and conclusions, and these appear to be sufficiently addressed.

Author Response

Please see attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report (Previous Reviewer 2)

The revised paper describes two high-altitude events observed by a lidar (as well as a dust layer in the boundary layer). One is said to be dust-related and the other not. Near the end of the paper, it appears that the authors believe that homogeneous nucleation is the most likely mechanism for the formation of cirrus, and that heterogeneous nucleation requires special conditions of dust layers. In fact, while INP are rare, there is almost always sufficient dust particles to allow heterogeneous nucleation of cirrus. Heavy aerosol loadings certainly have an impact, but the ‘natural’ level of dust does support the formation of cirrus. The efforts of the authors to separate dust from non-dust conditions leads to some extreme suggestions on the interpretation of observations.

The authors also seem to underestimate the impedance of vertical transport between the surface and the tropopause. In particular, they assume that high-altitude back-trajectories will reflect conditions at the surface. Without deep convection, surface pollution will be generally restricted to the boundary layer depth.

The English will require some work to clarify the meaning of some sentences.

 

Specific comments

l45-48. Twice in one paragraph it is stated that INP modifies heterogeneous nucleation. INPs are the basis of heterogeneous nucleation; the don’t ‘modify’ it. They modify the microphysical character of cloud.

L71 What is meant by ‘weak’ cold air?

L82. Is high precision more appropriate than high accuracy for lidar?

L83. Transmission → transport (this occurs on other occasions in the paper)

l162-177. The footprint of CALIOP is relevant.

L183. AOT is not defined.

L187. Bureau of Meteorology → Chinese Meteorological Agency

l190. Values of PM2.5 and PM10 are dependent on instrumentation; are the instruments known for CMA?

L278. Swap figures 2 and 3; figure 3 is discussed before figure 2.

Figure 3 has the captions swapped. Figure 3a shows relative humidity of 120% at temperatures around 220K where homogeneous ice nucleation would occur; is this relative to ice or water? It would be useful to show the sampling periods on this figure, as in figure 2.

The y-axis is adjusted pressure to correspond to height so that it can be compared with figure 2. It would be better if the coordinate was height itself to facilitate comparison.

L288. Surely at 223K the phase will be ice?

L358. Why would pollution lead to a reduction in aerosol height? The BLH usually changes due to dynamical forces; aerosol concentration then increases with decreasing BLH due to conservation of mass (unless the source also decreases). The ERA5 data suggest large-scale forces are changing the air mass over the period of interest.

Figure 4. The time of the satellite pass should be noted in the caption. The precise meaning of the red lines should also be in the caption.

L395. The ‘red frame’ lines need to be defined more precisely. Are they when the satellite path is within 100 km of the ground-based lidar?

L400. While the back-trajectory in figure 5a shows that the air came from regions with high surface pollution (table S1), the trajectory is always above 5 km. Is there any evidence of mixing from the surface to this height as the air moved eastward? It seems unlikely to me, unless the source sites are at very high altitudes. The back-trajectories should take into account convection.

L414. It seems an exaggeration to state that CALIPSO ‘proved that the second ice cloud was related to dust’. Figure 4 identifies a patch of dust on the edge of a large region of ice cloud, where ERA5 shows a large region of high-humidity air.

Figure 6 is very hard to interpret. Although Himawari8 has a resolution of a few kilometres, the figure shows images covering at least 500 km around the lidar. The relevant pixels are hard to identify. Moreover, I believe the optical depth is estimated over the whole atmospheric layer; not just the layers above 8 km.

L432. It is not correct to state that the Himawari8 data ‘proved’ that there was high-altitude aerosol. The height of the aerosol is not determined.

L444-464. Figure 7 is very confusing. The box plots show outlier values above the ‘upper limit’; what is the upper limit? It is also apparent that, while DC2 may be statistically different from C1..C4, there is no difference between DC1 and C1..C4. It should not take 20 lines of text to say this.

There also needs to be discussion of the expected difference in LDR between ice and aerosol. All the values shown in this paper seem to me to be within the range of normal cirrus cloud.

L478-485. As with figure 7, figure 8 shows images on a scale of about 600 km rather than showing a window near the lidar. Moreover the MODIS cloud top temperature seems to be much warmer than the cirrus at 8-10 km. Indeed the statement that the MODIS CTT is below 240K is strange when the lowest scale on the colour bar of figure 7 is at 240 K.

The MODIS effective radius is 48 and 32 um, which seems extremely large for aerosols; but quite likely for ice crystals.

L488. It is now suggested that in fact it is ice cloud that has been affected by aerosols. This is certainly likely. While INP are relatively rare, there is almost always sufficient for the nucleation of cirrus.

L500-505. It is now suggested that in the absence of dust layers, cirrus is formed by homogeneous nucleation. As noted earlier, while INP are rare, there is usually enough to nucleate cirrus before homogeneous nucleation occurs. If this is the author’s definition of dust layers, then dust is everywhere.

L547-550. This is a confused discussion of the Twomey effect, which applies to the nucleation of water droplets.

L553-562. The insistence on delineating dust vs non-dust cloud leads to complicated statements about small differences in depolarisation.

L563-569. In this paragraph, the authors state that the dust has no obvious effect on depolarisation, in contradiction to the previous paragraph.

 

Author Response

Please see attached file.

Author Response File: Author Response.pdf

Reviewer 3 Report (Previous Reviewer 3)

Please see attached file.

Comments for author File: Comments.pdf

Author Response

Please see attached file.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report (Previous Reviewer 3)

I am satisfied with corrections, it can go to publication now.

Author Response

We sincerely thank you for your suggestions and opinions all the time. These comments are all valuable and very helpful for revising and improving our manuscript, as well as the important guiding significance to our researches. 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

This is a good paper that helps to quantify lidar aerosol measurements with other types of instrumental sensors.   The data appears to be correct and shows the interplay of the backscatter and satellite measurements.   The data obtained shows the importance of this approach and should be useful for the atmospheric community.   One improvement that could be noted is to give the specifications for the lidar in greater detail such as pulse repetition frequency, exact wavelength/single frequency or linewidth ? , or a reference to the lidar parameters.   It is important to note that such a lidar design is eye-safe and could potentially be used for coordinated large scale field measurements.

Reviewer 2 Report

The paper describes a case study of a couple of high-aerosol events using a lidar. There is insufficient description of the instrumentation and of the synoptic situation and local geography associated with the observed high-aerosol events. The claimed new finding is of dust in high cloud, but the evidence needs to be strengthened considerably.

The lack of background information makes it very hard to interpret the results, and even to reconcile the lidar and PM10 measurements near the surface. The study could make much better use of Himawari data (with its 10-minute sampling).

Specific comments

l41-76. This is a long list of past projects using lidar: one sentence for each project with no apparent science link between projects. Are all these references needed?

L82-94. Following this list of earlier work and in the same paragraph is a description of the current work, followed by a reference to another earlier work. The next paragraph outlines the current paper. These sentences should be reorganised to clarify what is past work and what is the current work.

L102. Figure 1a uses a Google Maps background in Chinese, and so for many readers there is no idea of the geographical location of the instrument. There is no idea of the scale of the region. There should also be a description of the local geography. Is the region or rural? What are the local sources of aerosol.

L104-105. The ground-based meteorological variables have a ‘temporal resolution of 1 hour’. Are the recorded values averaged or instantaneous? What instruments were used to measure each variable? Where is the nearest upper-air station to provide vertical soundings?  (Note that the CALIPSO data were at least 100 km away.)

L108-122. The text and figure give an overview of the structure of the lidar. Is there a reference providing more detail on its engineering and performance, or a manufacturer and associated details? What is the response of the lidar to PM2.5 and PM10 particles?

L168-171. Reference 33 is available in Chinese: only the abstract is in English. As this paper describes the details of the technique, many readers will not be able to understand the method to solve for the boundary layer height. The three equations do not provide much insight. Is there a paper in English that describes the general approach and some evaluation of the method?

L178. The year of the observations should be included here.

L180-191. Before discussing the local conditions, it would be appropriate to describe the synoptic situation. Moreover, a sounding from the closest radiosonde site would also provide information on the synoptic situation and on the height of the ABL.

L197. “… when the PM10 rose rapidly, the boundary layer dropped.” I see no real evidence of the ABL reducing around 9am on 26 March from figure 2c.

l200-203.  When I look at conditions before and after 9am on 26 March near the surface, the lidar readings show little change; indeed the signals may be weaker after 9 am. I find it hard to reconcile the PM10 and lidar measurements near the surface.

L192-203. Is the change in aerosol associated with a synoptic or local change? The only real change seems to be a spike in PM10 and an increase in wind speed with a rise in temperature and a drop in humidity. What time is sun rise?

L206. What is the cloud at 8km due to a ‘thick mixture of aerosols and clouds’? The most likely explanation is high-level ice cloud associated with a synoptic change. Why would there suddenly be a high loading of aerosols at 8km?

L209. Which differences in lidar variables suggests that the signal at 8km is aerosol plus cloud, while the signal at 10km is cloud. They look the same to me; ie ice cloud.  

L217. The change in ABL depth at 1800 must be due to a synoptic change, especially with the change in high cloud. You need to look at the synoptic charts and satellite imagery to interpret these results.

Figure 3 caption. Give the actual time of the satellite data. It seems to me that the CALIPSO retrieval has the cloud between 8 and 20 km at the end of your period of interest, but your results have the cloud mainly at 10 km at that time. Does the nearest sounding help?

L238. Looking at the relative humidity and the backscatter in figure 3, it is hard to believe that the 10km signal is not just ice cloud. You need to consider the uncertainty in satellite retrievals.

L248-255. Himawari data are available every 10 minutes. Why do you not have a time series of AOT over your area for the full period?

L259-265. Which layer of the atmosphere does this discussion relate to? At high levels, there seems to be plenty of humidity, and at low levels it is probably too warm for ice and there is no indication of whether there is any cloud.

L294-302. You need to add error bars to each of the points in figure 5. The statistical significance of the difference must be demonstrated.

In order to have convincing evidence of substantial dust in the high cloud in figure 2, you need to show that the air mass arriving in the afternoon of 26 March could have a dust source.  What are the synoptic conditions?

 

 

 

 

 

Reviewer 3 Report

see attached document for corrections suggested and here are given some important points.

1. abstract/intro; objs are not clear and should be given in the end of initro.

2. obj should be clear in abstract

3. intro should be divided to focused points.

4. results; not clear some figs captions

5. some claims are baseless.

6. some results are not clear e.g., why smaller ice crystals will show large Dep Ratio?

7. Conclusions need to be itemized and related to be linked to the results

8. Refs are provided should be included and pointed for related text.

9. overall text flow needs to be improved (see comments on the paper).

Decision: Major revisions

Comments for author File: Comments.pdf