Air Quality Degradation by Mineral Dust over Beijing, Chengdu and Shanghai Chinese Megacities

Round 1

Reviewer 1 Report

Page 1, line 19-21. Shouldn’t this be particulate matter rather than particle matter? Also what size particle matter is referred to with the measure of 75ug/m3? And 772ug/m3? Is it pm10? Or pm2.5 or TSP?

 

Page 1&2- introduction 1st and 2nd paragraph, the cited PM concentrations do not mean much without the national air quality standards also cited. This would give more context to how high they are.

 

Page 2, line 58. What is the basis for the years selected for the modelling study?

 

Section 2.1 & 2.2 are extremely brief. Additional information should be provided about how all of the models cited work/are used to generate which aspects of the modelling analysis. Also, statements/justification regarding why the selected models are chosen should be included.

Author Response

The referee will find our full answer to review in the author coverletter (Please see the attachment).

Best regards

Author Response File: Author Response.docx

Reviewer 2 Report

This paper evaluates the dust contribution to PM10 concentrations at three Chinese cities, Beijing, Chengdu and Shanghai. The dust contributions are estimated via CHIMERE model simulations over the area, which, however, seem to have large discrepancies with MODIS AODs over the desert areas. So, since the simulated and modeled AODs are so different RMSE in the order of 60-150%, the uncertainty in the PM simulations at sites very far from these source regions is expected to be extremely high, since many other parameters, like local, regional, synoptic meteorology, mixing with urban-anthropogenic emissions, etc are involved. Indeed, the uncertainties in PM10 and PM2.5 estimates on hourly measurements are very high (Table 3). Note that these are the total PM levels, not the contribution of dust, as I can understand from the readers. So the contribution of simulated dust-PM to the total measured PM over these urban environment is highly uncertain and any results and discussions would be rather valueless. 

On the other hand, authors speak either for the three stations or for many stations and the reader is really confused of what he can expect from the current results. At the final stage, there is a figure providing also model simulations of PM chemical components, organic matter, BC, sulfate, nitrate, etc... How are they connected with dust contribution from the desert areas that is the objective of the current manuscript?

There is no chance for comparison of the CHIMERE and CALIPSO simulations... Referring to a figure in another manuscrpt and, especially to a climatological figure, is not the right way of presenting the results.

Many parts of the manuscript need strong revision. For example in lines 279-280 there is no verb in the two sentences.

Conrtributions of each desert are not clear in Fig. 1b, colors need to be changed. 

Overall, the literature is very rich with studies dealing with the dust pollution issue in Chinese megacities, especially during the spring season, synergizing ground measurements, chemical analysis, remote sensing and model simulations. Therefore, I cannot see the importance and the scientific contribution of this paper to the current knowledge in this issue. Also, authors do not clearly indicate/highlight the contribution of this manuscript. If this is the use of CHIMERE for the simulations of the PM10 and PM2.5 contributions of dust, then the uncertainties are so large that the current results cannot be considered as valid by the scientific community...That's just my opinion.  

Author Response

The referee will find our full answer to review in the author coverletter (Please see the attachment).

Best regards

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript studies the degradation of air air quality by mineral dust over China for 2011, 2013 and 2015 and authors performed model simulations. The results can be useful for PM pollution control in China. The manuscript is in general well written, but there is a need for English language editing. The topic of this paper is relevant to the journal Atmosphere.

Comments:

P2L56: Maybe you can break the sentence in two - the way how it is now it is not easy to understand.

P2L61: The English grammar for this sentence need to be checked.

P5L160: It’s difficult to understand these sentences.

P11L343: The English grammar need to be checked.

P13: It’s difficult to read Figure 5. Better presentation and interpretation are needed.

P15L455: The grammar for this sentence need to be checked.

In the conclusion it should be mentioned what the authors will focus on further research.

Author Response

The referee will find our full answer to review in the author coverletter (Please see the attachment).

Best regards

Author Response File: Author Response.docx

Reviewer 4 Report

This paper compares CHIMERE model results to local ground measurements and AOD data in 3 Chinese cities during 2011, 2013, and 2015. The paper's focus on the impacts of episodic dust events is useful and important.

However, additional clarifications of the text are necessary. Two important limitations to the work should be stated, in particular: 1) any incompleteness in the anthropogenic emission data indirectly affects the calculated seasonal percent contributions of windblown dust, and 2) since the model's PM2.5 and PM10 predictions have only been tested in rural and suburban areas, the validity of these predictions for urban areas within these cities is unknown.

 

20-22
Clarify whether these cited statistics represent PM2.5, PM10, or PMtotal.

25-26 and 30-31
Remove text from reference brackets and correct grammar so that these form complete sentences.

25-35
It appears that PM2.5 and NOx pollution in China has decreased 2007-2020, yet the next sentences state that sunshine is decreasing and health effects and industry are increasing. Please clarify what these contradictory trends are meant to communicate, particularly with respect to the goals of your study.

27-29
The disparate increases in premature mortality and population growth are difficult to evaluate as stated, and should be compared using similar units (e.g., % for both), or by comparing premature mortality/population ratios for 2000 vs. 2010.

97-98 and 335-338
Are the anthropogenic emission inventory inputs time resolved over 2011, 2013, and 2015, with seasonal inputs for each year? If not, any claims that the dust percent contribution is more variable than anthropogenic emissions could be partially an artifact of the model input datasets. Similarly, the uncertainty in anthropogenic contributions over time will contribute uncertainty to the percent dust estimations. Temporal meteorology variations in the model can account for anthropogenic variation to some extent, but cannot account for other seasonal variability such as factory shutdowns, construction seasons, or other episodic events.

138-144
Clarify why you chose a different PM2.5/PM10 dust mode parameterization than the cited references. Was this chosen empirically afterwards, to best fit the model to the ground measurements and AOD data?

158-160
Explain or cite in more detail how AOD was computed with CHIMERA.

201-206
State the PM measurement methods used for the ground monitoring data (e.g., filter-based, beta attenuation, or optical).

209-13
The ineligibility of measurement stations near sources implies that the CHIMERE model performance cannot be verified for urban or traffic locations. This represents a limitation of this work for comparing dust impacts to anthropogenic sources. Specifically, the accuracy of the model-predicted dust impacts in these 3 cities are only known for suburban and rural locations within these cities. This limitation should be stated here and in the overall conclusions.

Table 2, 264
Define NRMSE and RMSE.

266-277, Fig S1, Table S1
Given that AMPP is a substantial portion of the predicted PM10 and PM2.5, it seems possible that a portion of the AOD signal is affected by anthropogenic dust in addition to natural dust. Does CHIMERE account for AMPP in its AOD predictions? If not, this might be a source of the discrepancy between MODIS and CHIMERE.

278
"Similar" would be more appropriate here than "really close".

302-304
Please clarify this sentence or delete.

338-339
More explanation is needed for the claim that decreasing PM10 2011-2015 is due to changing anthropogenic contributions.

Author Response

The referee will find our full answer to review in the author coverletter (Please see the attachment).

Best regards

Author Response File: Author Response.docx

Reviewer 5 Report

The authors here are presenting their efforts to quantify the dust contribution on PM10 over various megacities in China. Towards this objective they use data obtained by real measurements from ground and space as well as modelled concentration. The manuscript is well written and holds a scientific merit. Therefore, in my opinion it can be published as it is in the Journal Atmosphere of MDPI.

 

The authors are demonstrating their efforts towards quantifying the contribution of dust in air quality over three megacities in China. In order to do so they use a dataset composed by in situ and space based observations as well as numerical simulations. The manuscript is well written and has a scientific merit. In my opinion it can be published in the journal of Atmosphere. In order to be improved I would kindly suggest the authors to follow some comments below:

1. Line 60: “AOD” -> “Aerosol Optical Depth (AOD)”
2. Lines 205-206: A short text is missing here to describe a bit how accurate these data are and from which instruments were obtained.
3. Line 266: 1 x 1 deg resolution is too coarse. The authors have to justify why they used AOD from MODIS and not from other spaceborne instrument with finer spatial resolution.
4. The authors are kindly requested also to elaborate more on the reasons behind the observed differences between MODIS and CHIMERE AOD retrievals
5. Line 309: Please rephrase
6. Line 476: “ane xeedance”->“an exceedance”

Author Response

The referee will find our full answer to review in the author coverletter (Please see the attachment).

Best regards

Author Response File: Author Response.docx

Reviewer 6 Report

The manuscript titled “Air quality degradation by mineral dust over Beijing, Chengdu and Shanghai Chinese megacities” examines contributions of dust aerosol to the particular matter (PM) mass concentration found in three cities in China based primarily on the numerical modeling for three years, 2011, 2013, and 2015. The authors conclude that dust is one of the most important aerosols that are responsible for the high PM mass concentration for the three cities during spring when the dust emission is more frequently to occur. Additional analyses are needed to make this study complete and the results solid. The writing is in general good, still improvements are required, but a few sentences are unclear. I report some but not all of them in the following specific comments. Great efforts are needed to improve the overall quality of the manuscript before acceptance for publication. I suggest major revisions for further consideration.

My main concerns:

• A selection of the three years by the authors results in a representative issue. As stated by the authors, “numerous dust events have been observed” during these three years. Conclusions based on these years may not be a representative of the contribution of dust aerosol to the total PM mass concentration in the three cities for normal years.
• To achieve the authors’ goal, one of the key things are that if the model can reproduce AOD of each component well, compared to observations such as MODIS retrievals, and PM observations. This validation targeting at each component is missing.
• In addition to averaged metrics for the model performance evaluation over the sub domains, it would be also required to show if the model can reproduce the spatial distribution of AOD from MODIS.
• For the dust vertical dispersion, the results shown in this study are not from the climatology perspective but for specific years, so the authors should be very careful when comparing the dust vertical dispersion from others. I would suggest to compare the simulated extinction coefficient profile to CALIPSO observations.

Specific comments:

Line 5-6: Consider change to “we quantify the degradation of air quality by dust over…using the chemistry transport model CHIMERE, which simulates dust emission and transport online”

Line 6: Spell out CHIMERE

Line 37: The range of the global dust emission estimate that the authors shows is too narrow. Please see Hamilton et al. (2019).

Hamilton, D.S., Scanza, R.A., Feng, Y., Guinness, J., Kok, J.F., Li, L., Liu, X., Rathod, S.D., Wan, J.S., Wu, M. and Mahowald, N.M., 2019. Improved methodologies for Earth system modelling of atmospheric soluble iron and observation comparisons using the Mechanism of Intermediate complexity for Modelling Iron (MIMI v1. 0). Geoscientific Model Development (Online), 12(9).

Line 39: Dust aerosol is important not only because of its mass, but also its direct and indirect effects: for example, mineral dust can perturb the radiative energy budget of the Earth system.

1. Sokolik, I. N., & Toon, O. B. (1999). Incorporation of mineralogical composition into models of the radiative properties of mineral aerosol from UV to IR wavelengths. Journal of Geophysical Research: Atmospheres, 104(D8), 9423-9444.
2. Li, L., & Sokolik, I. N. (2018). The Dust Direct Radiative Impact and Its Sensitivity to the Land Surface State and Key Minerals in the WRF-Chem-DuMo Model: A Case Study of Dust Storms in Central Asia. Journal of Geophysical Research: Atmospheres, 123(9), 4564-4582.

Line 43: to be 252...

Line 47: Change “yearly” to “annual”

Line 49-50: Why do the authors show specific results of this study in the introduction section?

Line 63-64: Need to rewrite this sentence, “It has also been possible to determine …”

Line 99: A little more information is needed about the “remote sensing observation”

Line 118: Delete “in”

Line 135: Are all these three parameterizations used in this study? Only need to mention the one you are using

Line 141: “of” to “from”

Line 155: Specify which satellite. One around 10:30AM, the other around 1:30 PM

Line 193: What kind of interaction do you mean?

Line 196: Locations of the observation stations are required

Table2: Please explain r and NRMS. Do the same thing for other tables if needed.

Line 415: Are these results for 2013?

Line 427: What's the meaning of "very localized in time? Shouldn't dust particles in these cities are transported from remote dust sources?

Figure 5: which year? All three years? Also, I did not see a), b), and c) in the figure.

Figure 6: “[00;10[“ to “[00; 10]”, and many others.

Author Response

The referee will find our full answer to review in the author coverletter (Please see the attachment).

Best regards

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The revised manuscript has been significantly improved in its structure, literature overview and scientific discussions of the main results. Authors followed reviewer comments and also provided a clear response to all comments. The final manuscript has to be checked in the proofs procedure for some gramatical errors like "Dust outbreaks largely contributes" (end of abstract) and some others in the main text. 

 

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 6 Report

I am wondering again whether other components are well reproduced in addition to dust compared to observations. No information regarding evaluation of the non-dust aerosols is shown in the manuscript. Therefore, I am not convinced by the contributions the authors estimated by dust to the total PM10. The uncertainty in this estimation should be discussed at least in the text.

Line 50: The reference [15] the authors cite does not well support the statement. Please see my report in Round 1.

Line 195-200: Is this an apple-to-apple comparison? As the satellite overpasses in the morning, the model result should also be for a time around 09:30 AM.

Line 204-207: I don’t think there is a result-cause logic between the two.

Line 345: What do you mean by saying “which is a wavelength characteristic of dust”? I think just saying “at the wavelength of 10 um” should be fair.

Line 348-355: You are using three-year simulations to calculate the dust frequency. I don’t think it is statistically significant. This should be pointed out.

Line 348-349: Delete “which lined to…”.  It is irrelevant providing no useful information.

Author Response

please see the attachment

Author Response File: Author Response.docx