Evaluating the Performance of MODIS and MERRA-2 AOD Retrievals Using AERONET Observations in the Dust Belt Region

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The abstract needs to be completely rewritten and the results need to be written in a more scientific manner in the abstract. The introduction is very long and much of the information presented is irrelevant to the topic of the MS. The way the figures are presented is not appropriate. It is necessary to put the results for each station from different data sets together to make them easier to compare. The graphs of some figures can be shown on one figure to reduce the number of figures and make the comparison easier. Some figures should be moved to the supplementary section. For some figures and tables, very long explanations were written, and others were not analyzed at all. Throughout the text and conclusions, the results were not compared with previous studies, which are also a considerable number. Overall, the article needs a lot of corrections. See the attached file for other corrections and details.

Comments for author File: Comments.pdf

Author Response

Comment 01:  The abstract needs to be rewritten, its structure is not appropriate. The method needs to be written coherently in a few lines and then the most important results presented. The results written are too general, the results should be presented in a more quantitative and precise way. In general, the abstract needs to be written in a more scientific way.

Reply 01:         As per advice of the learned referee the “Abstract” has been rewritten by incorporating the suggestion in the revised manuscript.

 

Comment 02:  Since this part is not studied in this manuscript, it needs reference and should not be mentioned in the abstract.

Reply 02:         The said part has been removed from the revised manuscript.

 

Comment 03:  The introduction is very long and much of it has nothing to do with the topic of the manuscript. The paragraphs are also very long, which makes the reader tired. It is necessary to refer to similar works and their results in the introduction after a few general lines.

Reply 03:         We have reduced the introduction section as per advice of the learned referee in the revised manuscript. Hope this will be fine.

 

Comment 04:  This entire section is about dust sources, which is unrelated to the topic of the manuscript.

Reply 04:         The said section has been removed from the revised manuscript.

 

Comment 05:  Please move this part to the methodology section.

Reply 04:         The said part has been moved to the methodology section.

 

Comment 05:  This part is not related to this section.

Reply 05:         The part has been deleted from the revised section.

 

Comment 06:  it needs reference.

Reply 06:         Reference has been added in the revised manuscript.

 

Comment 07:  reference

Reply 07:         Reference has been added in the revised manuscript.

 

Comment 08:  reference

Reply   08:       Reference has been added in the revised manuscript.

 

Comment 09:  Summarize this section and separate the data and methodology sections. It is also best to have a separate subsection for each dataset.

Reply 09:         As advised, the data section has been separated from methodology section. In addition, subsection has also been included and the entire section is summarized in the revised manuscript.

 

Comment 10:  It is better to move this table to the next section that is mentioned in the text.

Reply 10:         Table 1 has been moved in the “Data” section.

 

Comment 11:  It is better to add number at the title of these columns.

Reply 11:         The table has been amended as required in the revised manuscript.

 

Comment 12:  The description of each figure or table should be provided directly above it. Throughout the manuscript, a series of consecutive contents are written, followed by several figures without any text.

Reply 12:         Changes have been made as per the suggestion of the learned referee.

 

Comment 13:  No explanation is provided in the text regarding Figures 2 to 4. It would be better to present all data at one station in one figure and compare them, then move the data for other stations to the supplementary section.

 

Reply 13:         As per suggestions, one figure has been added in the main text while remaining data has been moved to supplementary section of the revised manuscript. In addition, explanation of the amended figure has been provided. We hope this will satisfy the learned referee.

 

Comment 14:  Summarize the explanation for Table 2. It is not necessary to write all the numbers in the text, but state the important results and their possible reasons. It is also necessary to compare and analyze your results with other studies conducted in each region.

Reply 14:         Suggested changes have been made as per learned referee in the revised manuscript.

 

Comment 15:  Since the purpose is the comparison of AOD from different data sets, it is better to display the figures for the AOD annual and seasonal averages from different datasets together so that they can be compared better. For example, the AOD average in the spring season from different datasets should be shown in one figure and the other seasons in the same way.

Reply 15:         Thank you very much for nice suggestion. As per desire the changes have been made in the revised manuscript.

 

Comment 16:  Display all annual average of AOD charts from different data sets on one chart for each station. In this case, Figures 8 to 10 will be presented as one figure. Compare the trend of AOD changes with previous studies in each region.

Reply 16:         The suggested changes have been made in the revised manuscript. We hope this will satisfy the learned referee.

 

Comment 17:  Show the graphs of Figures 11 to 13 in one figure as well.

Reply 17:         The changes have been made in the revised manuscript.

 

Comment 18:  The time period should be mentioned in the caption.

Reply 18:         The time period for each AERONET station has been added in the revised manuscript.

 

Comment 19:  Figures 11 to 13 do not show seasonal trend. They just show seasonal average in the time period. It is better to summarize the explanation.

Reply 19:         The required changes have been made in the revised manuscript.

 

Comment 20:  reference.

Reply 20:         References have been added.

 

Comment 21:  This part is not appropriate for this section.

Reply 21:         As per suggestion of the learned referee the entire paragraph has been removed from the revised manuscript.

 

Comment 22:  This issue is not discussed in the text.

Reply 22:         Yes that is correct and therefore the said text has been removed from the revised manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

Review of Manuscript: No. earth-3642053

This study used AOD data from 16 AERONET sites to evaluate the performance of MERRA-2, MODIS Aqua, and Terra AOD products, and revealed the spatiotemporal variations in the retrieval accuracy of MERRA-2, MODIS Aqua, and Terra AOD. Based on the experimental results and technical approach, the study is considered reasonable. 

Major Comments:

- The authors mentioned in both the Introduction and Study Area sections that the selected AERONET sites represent arid, semi-arid, coastal, and urban environments, covering urbanized areas, arid regions, coastal zones, and biomass burning hotspots. In the results section, the authors also provided a detailed analysis of the spatiotemporal variations of AOD at each site. However, they did not summarize the performance of the three AOD datasets in these specific types of environments. It is recommended that the authors include a summary of how the datasets perform across these environmental categories.

- When analyzing the spatiotemporal variation of AOD at the Lahore site, the authors attributed part of the reason for the high AOD load in winter to the inflow of pollutants from China. As far as I know, in addition to the crop residue burning in rural areas and frequent fog and smog already mentioned by the authors, vehicle exhaust emissions are also likely to have an impact. Moreover, the presence of inversion layers makes it difficult for high concentrations of AOD to disperse, resulting in peaks during the winter. In addition, the region most likely to influence AOD in Lahore should be northwestern China. However, to my knowledge, industrial pollution in this region is mainly concentrated in basins, and the terrain to some extent limits the dispersion of pollutants. Furthermore, only dust is capable of relatively stable long-distance transport, which mostly occurs in the spring. Therefore, attributing the high AOD load in Lahore to the diffusion of pollutants from China may be somewhat far-fetched.

- It is suggested that the author could consider more technical methods, such as the HYSPLIT, when analyzing the sources of pollution. In addition, it is also recommended that the author refer to the following researches to enhance the understanding of the regional aerosol and highlight the significance of this study in the introduction section：

https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2024GL113448

https://doi.org/10.1016/j.rse.2021.112410

 

Minor Comments:

- Please pay attention to the formatting: line numbers appear above Figure 2 and the subsequent figures and tables.

- Figures 2, 3, 4: Is it necessary to directly display the R² values within the figures to facilitate more intuitive comparison and analysis?

- Figure 5, 6, 7: Please enlarge the text for better readability.

- Figure 5, 6, 7: What do the numerical values labeled on the grid in the figure represent? They do not seem to be longitude and latitude. 

- Figure 8, 9, 10: Merge the three figures according to their corresponding sites into a single figure to facilitate comparison and analysis. Figures 11, 12, and 13 should be handled similarly. 

- Figure 8: The dataset evaluated is MERRA-2, not MERRA. Figure 11 is similar. 

Author Response

Comment 01:  This study used AOD data from 16 AERONET sites to evaluate the performance of MERRA-2, MODIS Aqua, and Terra AOD products, and revealed the spatiotemporal variations in the retrieval accuracy of MERRA-2, MODIS Aqua, and Terra AOD. Based on the experimental results and technical approach, the study is considered reasonable.

Reply 01:         Thank you very much for your thoughtful comment. We truly appreciate your recognition of the methodology and findings, as well as your acknowledgment of the study’s reasonableness. Your feedback is encouraging and valuable.

 

Comment 02:  The authors mentioned in both the Introduction and Study Area sections that the selected AERONET sites represent arid, semi-arid, coastal, and urban environments, covering urbanized areas, arid regions, coastal zones, and biomass burning hotspots. In the results section, the authors also provided a detailed analysis of the spatiotemporal variations of AOD at each site. However, they did not summarize the performance of the three AOD datasets in these specific types of environments. It is recommended that the authors include a summary of how the datasets perform across these environmental categories.

Reply 02:         The about the performance of these datasets is available in the “Conclusion” section. However, if learned referee wants us to elaborate it, we are happy to do it.

 

Comment 03:  When analyzing the spatiotemporal variation of AOD at the Lahore site, the authors attributed part of the reason for the high AOD load in winter to the inflow of pollutants from China. As far as I know, in addition to the crop residue burning in rural areas and frequent fog and smog already mentioned by the authors, vehicle exhaust emissions are also likely to have an impact. Moreover, the presence of inversion layers makes it difficult for high concentrations of AOD to disperse, resulting in peaks during the winter. In addition, the region most likely to influence AOD in Lahore should be northwestern China. However, to my knowledge, industrial pollution in this region is mainly concentrated in basins, and the terrain to some extent limits the dispersion of pollutants. Furthermore, only dust is capable of relatively stable long-distance transport, which mostly occurs in the spring. Therefore, attributing the high AOD load in Lahore to the diffusion of pollutants from China may be somewhat far-fetched.

Reply 03:         Thank you very much for your valuable input. We have revised the text accordingly and hope it now meets the expectations of the esteemed referee.

 

Comment 04:  It is suggested that the author could consider more technical methods, such as the HYSPLIT, when analyzing the sources of pollution. In addition, it is also recommended that the author refer to the following researches to enhance the understanding of the regional aerosol and highlight the significance of this study in the introduction section：

https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2024GL113448

https://doi.org/10.1016/j.rse.2021.112410

Reply 04:         We sincerely thank the referee for this valuable suggestion. While we agree that the use of more advanced trajectory models such as HYSPLIT could provide deeper insights, incorporating such methods would significantly alter the scope and direction of the present study. Since our focus here is on AOD climatology, we have not included HYSPLIT analysis at this stage. Nevertheless, we fully acknowledge the importance of this approach and consider it a promising avenue for future research.” The said paper has also been added in the revised manuscript.

 

Comment 05:  Please pay attention to the formatting: line numbers appear above Figure 2 and the subsequent figures and tables.

Reply 05:         We have done this in the revised manuscript and hope it is fine now.

 

Comment 06:  Figures 2, 3, 4: Is it necessary to directly display the R² values within the figures to facilitate more intuitive comparison and analysis?

Reply 06:         After incorporating the suggestions of other reviewers we have moved scatter plots in the supplementary section and therefore R² values are required to be in the table.

 

Comment 07:  Figure 5, 6, 7: Please enlarge the text for better readability.

Reply 07:         Thank you very much for pointing out the matter. Although, we have increased the font but the final formatting will be done once the manuscript is accepted for publication.

 

Comment08:   Figure 5, 6, 7: What do the numerical values labeled on the grid in the figure represent? They do not seem to be longitude and latitude.

Reply 08:         Indeed they are longitude and latitude.

 

 Comment 09: Figure 8, 9, 10: Merge the three figures according to their corresponding sites into a single figure to facilitate comparison and analysis. Figures 11, 12, and 13 should be handled similarly.

Reply 09:         The changes have been made in the revised manuscript as suggested by the learned referee.

 

Comment 10:  Figure 8: The dataset evaluated is MERRA-2, not MERRA. Figure 11 is similar.

Reply 10:         As many of the figures are changed we hope this typo would not be existed in the revised manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The paper "Evaluating the performance of MODIS and MERRA-2 AOD retrievals using AERONET observations" presents the results of aerosol optical depth in the "dust belt" region. The results offer a very rough idea of the aerosol optical characteristics in the region since only the AOD is presented.

The following are major comments of the paper as a whole:

1. The paper uses AERONET measurements from 10 sites. Yet, only two authors are written in the paper. I suggest that the authors acknowledges their names or consider them as authors (based on the fact that they are responsible for data gathering). They can be consulted about weather or climate activities to further make the discussion on the sources and temporal changes better and more specific. If one reads the results, which just enumerate the results and provide general reasons of aerosol sources, one can not get new insights and scientific information.
2. The authors have presented the RMSE, MAE, RMB, SDA and SDM results from each site. In the discussion section, there is no mention on the significance of these values and why these values are important. In some locations, the values are almost similar. What would be the underlying reasons?
3. Why are the AERONET data not obtained from the AERONET website? The manuscript does not even say what AERONET version and AOD level are used.
4. The analysis of AOD is somewhat roughly done. The statistics of the data at each site can provide information on how the standard deviation is changing when one looks at the temporal (or seasonal) analysis.
5. The authors have not provided literature on the sources of aerosols in each site. A general description is just written and therefore this does not provide good support of the result. I believe this where the principal investigator can be of help.
6. The manuscript lacks the information of the referenced climatological information of the sites. This is important to have ideas when aerosols are affected by relative humidity. In the discussion, the authors should discuss by how much are the aerosol AOD in each site are affected by relative humidity. The authors should note that aerosol growth (due to relative humidity) increases aerosol AOD and does not indicate more aerosols. How much of this phenomenon is present in each site? I believe that the answer to this question in the manuscript will provide new findings rather than just enumerating the results and use general climatological description to support the findings.

Other comments:

1. The abstract is somewhat weak. The connection between the bigger problem vs what is done is not smoothly connected. The abstract does not indicate accuracy values (which can be considered something new if the approach presented in the paper has novelty). Stating that comparison is done in the study is not sufficient. The authors must state which parameters compared give good results and explain why (in the text).
2. L 57-61: This part is somewhat misleading. Not all aerosols in the "dust belt" region are dust (as seen in the result). Yet, the authors put too much importance on this. Why? How much (quantitatively asking) of the results are from dust if the reader puts importance to this paragraph?
3. L. 81-82: Can the authors provide references of this claim?
4. L. 84: The sentence is a very misleading. Not all satellites are have polar orbits. Geosynchronous satellites exist and the data from these type of satellites are very important and can also observe dust in real time.
5. L. 88: The statement contradicts with the second statement in the paragraph.
6. L. 99-101: Misleading statements. Ground instruments have limitations too. If there are clouds and dust event occur, sun photometer data are useless. Consider also the fact that sun photometers can't provide data at night even during dust events.
7. L. 145-147: Judging from the flow of thought (discussing dust to discussing aerosols), the authors have to state here that AOD of aerosols are not necessarily AOD from dust.
8. L. 157-162: Why is the effect of relative humidity in these observations not stated? How do relative humidity changes the AOD in each site. The authors must take into consideration that high AOD due to high RH does not mean high AOD from dust. Without this distinction, the discussion of the results presented in the manuscript is highly questionable! 
9. L. 233-236: Giovanni website is used only to visualize the data before serious analysis is done and NOT for downloading data. The Giovanni website is used to aid researchers not waste time working on large data set not knowing what results they should be expecting. After knowing the expected result (which is a rough picture presented in Giovanni website), data should be downloaded from respective archive, e.g., LAADS DAAC.
10. L. 252-253: Better use the word "level" and indicate the level number so as to be consistent with how AERONET data are categorized.
11. L. 256: atmospheric clarity=> weird choice of word.  
12. L. 259-260: What AERONET version and AOD level is used? The authors indicated that the AERONET data is retrieved from the AERONET website but the acknowledgment section contradicts this. Which is the true case?
13. L. 290-291: Again what AERONET level is used?
14. Figs. 2, 3 and 4: The authors did provided the state and end dates of the data presented in the graph, just the number of points (Table 1). Do the data have significant points for temporal analysis? Are there gaps (temporal) in the data that the reader should be aware of? What are the equations and r² values of each graph?
15. L. 455 & 458: AERONET => Terra?
16. L. 470-471: The Angstrom coefficient from AERONET data should provide information on the presence of fine and coarse (dust) aerosols in the overall result rather than saying flatly "likey due to local pollution and complex...". This is missing in the analysis and the reader is just bombarded by generalities when specific information can be explicitly stated to increase scientific knowledge.
17. L. 492-494: Really? If this is the case, there should be some systematic error since the same algorithm is applied in retrieving the aerosol optical properties from the data from each site. 
18. L. 505-508: Should these results verify the results in Sec. 4.1?
19. L. 509-514 (and also for other similar paragraphs describing the datasets from other sites): The information of the dataset's time, month , year is not stated. The authors should note that the data presented here includes the effects of relative humidity which increases AOD but does not mean "more aerosols".
20. L. 531-533: So what are in these sites that produce different results compared to other sites? Is it due to more aerosol sources, aerosol growth or dust (which can be confirmed from AERONET data)?
21. L. 533-534: This is a misleading/ambiguous statement. What statistical parameters do the authors refer to?
22. Figs. 5, 6 and 7: The colorbar is missing. Are the colors in the each figure have the same minimum and maximum values? If not, then the discussion has to be re-evaluated. Which of the graphs are a), b), c), etc.? What are the years that represent these data? What are the standard deviation graphs?
23. L. 681-682: Theoretically, what values of optical properties (e.g., AOD) are to be expected under these type of environments? I think the reader has to know this information so he can evaluate the theoretical and the results presented in this paper.
24. L. 692-699: A short description of each site may help the reader understand the results.
25. L. 694-695: References to this claim?
26. L. 713: References to this claim?
27. Figs. 8, 9, 10: Whey is MERRA AOD higher than AERONET? Are the mean values within the standard deviation of each measurement point? Why is the standard deviation not shown? Why display annual result when the authors' objective is to know how MODIS and MERRA measurements differ from AERONET measurements.  Why not compare daily dataset and study the difference of these daily differences from these measurements?
28. Figs. 11, 12 and 13: The wavelength information of the AOD is not stated. Which wavelength is used? The results in the graphs seem to contradict the results in the previous figures (Figs. 8, 9 and 10) that show that AOD from MERRA is greater than the AOD from AERONET.
29. L. 979 (and all related statements describing the AOD at each site): What are the standard deviation values? Are the mean AOD values at each site presented here show significant difference from the mean AOD values from other sites? Why?
30. L. 1003-1004 (and all related statements describing the years with high and low AOD): Why are the years of maximum and minimum AOD values different? When instruments gather data from the same parameter, the results should show the same trend of high and low values. There maybe differences but the trend should be the same. 
31. L. 1008: Indicate/define the "spring months" here.
32. L. 1015-1017 (all related statements describing the AOD values from AERONET and MERRA and MODIS): The positioning of the graphs is does not follow the narrative. The authors may need to rearrange the graphs to fit the narrative in the text.
33. L. 1027: How hot? The authors need to have specific description.
34. L. 1038-1040: References (ground-based measurements) to support this statement is needed.
35. L. 1058-1059: Quantitative values of temperature and humidity are not stated. How are these temperatures compare to other sites. The reader does not have an idea of how these temperature and humidity values differ from other sites.
36. L. 1060-1061: The authors state the same reasons in other sites. Which site then does not have secondary aerosol formation under the same season? Why?
37. L. 1076: What geographical characteristics? The authors need to be specific in this description. If the authors collaborated with the PI of each site, they would have a better knowledge of the site's climate/weather characteristics and may be weather data to support the authors' claims in the text.
38. L. 1077:  "prone to dust activity..." => So does other sites as mentioned in earlier paragraphs. Do the AERONET data support this claim?
39. L. 1080: "local farming practices..." => Not existing on other sites? How much of this activity is present in each site? Are there references to support the authors' claim?
40. L. 1084-1085: "regional influences" => references?
41. L. 1100: "dust activity" => This can be verified from the AERONET data. Does the AERONET data support this claim by the authors?
42. L. 1104-1105: These descriptions (and also of other sites) are too general. The authors should provide references to support these descriptions and claims.
43. L. 1112: "year 2015" => Why 2015? What's the cause?
44. L. 1114-1117: The authors need to provide references to support this claim.
45. L. 1123-1126: These descriptions are too general and does not directly give scientific proof.
46. Conclusion: The conclusion can be improved after major revision of the paper. The authors should directly point out if the objectives are achieved and respond to the the question "Are the results scientifically significant?"

Comments on the Quality of English Language

The quality of the usage of the English language can be improved. The authors should not use misleading and ambiguous words or sentences.

Author Response

Comment 01:  The paper "Evaluating the performance of MODIS and MERRA-2 AOD retrievals using AERONET observations" presents the results of aerosol optical depth in the "dust belt" region. The results offer a very rough idea of the aerosol optical characteristics in the region since only the AOD is presented.

Reply 01:         We respectfully disagree with this observation. The focus of the current study is specifically on evaluating the performance of MODIS and MERRA-2 AOD retrievals against AERONET observations, with an emphasis on spatial and temporal trends of aerosol optical depth (AOD) in the "dust belt" region. While we acknowledge that aerosol optical characteristics (such as single scattering albedo, Angstrom exponent, and refractive index) provide additional insight into aerosol type and behavior, these represent a distinct area of research that goes beyond the scope and objectives of this study. Our intention was to maintain a focused analysis on AOD retrieval performance, and we consider this a necessary and foundational step for future investigations into more detailed aerosol characteristics.

 

Comment 02:  The paper uses AERONET measurements from 10 sites. Yet, only two authors are written in the paper. I suggest that the authors acknowledges their names or consider them as authors (based on the fact that they are responsible for data gathering). They can be consulted about weather or climate activities to further make the discussion on the sources and temporal changes better and more specific. If one reads the results, which just enumerate the results and provide general reasons of aerosol sources, one can not get new insights and scientific information.

Reply 02:         We strongly disagree with the assumptions made in this comment. The data used in this study—AERONET Level 2.0 aerosol optical depth observations—are publicly available through NASA's AERONET database and are provided precisely for the purpose of scientific analysis by the broader research community. As per standard academic practice, proper citation of the data source has been provided. Including the principal investigators or data managers of the AERONET sites as co-authors is neither customary nor ethically required, unless they were directly involved in the design, analysis, or writing of the study.

The suggestion that our analysis lacks scientific merit or insight because it does not include these individuals as co-authors or advisors is unfounded and dismissive of the substantial effort involved in the research. Our interpretation of aerosol sources and temporal changes is based on well-documented meteorological patterns and satellite-based retrievals, and while we welcome constructive suggestions for improving depth, implying a lack of originality or scientific value is not a fair or evidence-based critique.

We respectfully remind the reviewer that scientific contributions should be evaluated on the rigor of methodology and clarity of analysis—not on the number of authors or speculative assumptions about external involvement.

 

Comment 03:  The authors have presented the RMSE, MAE, RMB, SDA and SDM results from each site. In the discussion section, there is no mention on the significance of these values and why these values are important. In some locations, the values are almost similar. What would be the underlying reasons?

Reply 03:         We appreciate the reviewer’s attention to the statistical metrics used in our analysis. However, we would like to clarify that metrics such as RMSE, MAE, and RMB are standard and widely accepted tools for evaluating model performance and retrieval accuracy, particularly in satellite validation studies. Their definitions, interpretations, and relevance are well established in the remote sensing and atmospheric science literature.

Given the suggestions from other reviewers to condense the manuscript, we opted not to elaborate on the basic significance of these statistical indicators in the discussion section in order to maintain focus on site-specific performance trends and comparative evaluation. That said, we have ensured that all metric results are clearly presented and interpreted where necessary. Minor variations or similarities in these values across sites reflect both consistent retrieval quality in certain conditions and site-specific influences, which are briefly touched upon in the discussion.

Should space permit or upon editorial request, we are willing to expand further on the contextual interpretation of these values in the revised version.

 

Comment 04:  Why are the AERONET data not obtained from the AERONET website? The manuscript does not even say what AERONET version and AOD level are used.

Reply 04:         We believe there may be a misunderstanding here. The AERONET data used in our study were indeed obtained directly from the official AERONET website (https://aeronet.gsfc.nasa.gov/), as clearly stated in the Data and Methodology section of the manuscript. Additionally, we specified that Level 2.0 AOD data—representing quality-assured and cloud-screened measurements—were used in our analysis. If this information was not sufficiently emphasized or clearly located, we will ensure that the statement is made more prominent in the revised version to avoid any ambiguity.

 

Comment 05:  The analysis of AOD is somewhat roughly done. The statistics of the data at each site can provide information on how the standard deviation is changing when one looks at the temporal (or seasonal) analysis.

Reply 05:         We find this comment somewhat unclear and difficult to interpret in its current form. None of the other reviewers raised similar concerns, and the statement lacks specific direction or actionable suggestions. It is not evident whether the reviewer is referring to a need for additional statistical analysis, or simply suggesting a different presentation. We would also like to emphasize that the temporal analysis—including seasonal breakdowns and associated variability—was carefully conducted and is reflected in both the visual and statistical summaries in the manuscript.

 

Comment 06:  The authors have not provided literature on the sources of aerosols in each site. A general description is just written and therefore this does not provide good support of the result. I believe this where the principal investigator can be of help.

Reply 06:         We respectfully find this comment to be both unclear and inappropriate in its suggestion. It is not evident why the reviewer believes that the Principal Investigators (PIs) of the AERONET sites—who are not involved in this research—should be consulted or included as co-authors. The PIs’ role is limited to the management and provision of ground-based measurements, which are publicly available for independent scientific use through the AERONET database. As stated earlier, appropriate citation has been provided in accordance with AERONET data usage policy.

Regarding the literature on aerosol sources, we provided general—but regionally grounded—explanations supported by established meteorological knowledge and previous studies. We opted for a broader discussion rather than repeating localized source attribution for each site, to maintain clarity and focus. If the reviewer has specific sources or site-related studies in mind that could enrich the discussion, we are open to considering them, but the current suggestion as phrased lacks clarity and scientific grounding.

 

Comment 07:  The manuscript lacks the information of the referenced climatological information of the sites. This is important to have ideas when aerosols are affected by relative humidity. In the discussion, the authors should discuss by how much are the aerosol AOD in each site are affected by relative humidity. The authors should note that aerosol growth (due to relative humidity) increases aerosol AOD and does not indicate more aerosols. How much of this phenomenon is present in each site? I believe that the answer to this question in the manuscript will provide new findings rather than just enumerating the results and use general climatological description to support the findings.

Reply 07:         We appreciate the reviewer’s point regarding the potential influence of relative humidity (RH) on aerosol optical depth. However, we would like to emphasize that the primary objective of this study is to evaluate the performance of MODIS and MERRA-2 AOD products against AERONET observations—specifically focusing on retrieval accuracy and spatial consistency—rather than investigating the physical processes that modulate aerosol properties at each site.

We acknowledge that RH can influence aerosol hygroscopic growth and thus impact AOD measurements. While this is an important factor in aerosol-climate interactions, quantifying the contribution of RH at each site would require either (a) integrating collocated RH data at high temporal resolution, or (b) conducting a dedicated process-level study—both of which are beyond the scope and intention of the current paper. Our use of general climatological descriptions was a deliberate choice to provide environmental context without diverting from the paper’s core validation goal.

Nonetheless, we will consider briefly mentioning this limitation and the potential influence of RH on AOD in the revised discussion, as a point for future study.

 

Comment 08:  The abstract is somewhat weak. The connection between the bigger problem vs what is done is not smoothly connected. The abstract does not indicate accuracy values (which can be considered something new if the approach presented in the paper has novelty). Stating that comparison is done in the study is not sufficient. The authors must state which parameters compared give good results and explain why (in the text).

Reply 08:         The abstract has been modified.        

 

Comment 09:  L 57-61: This part is somewhat misleading. Not all aerosols in the "dust belt" region are dust (as seen in the result). Yet, the authors put too much importance on this. Why? How much (quantitatively asking) of the results are from dust if the reader puts importance to this paragraph?

Reply 09:         We respectfully disagree with the concern raised. The focus of our study is explicitly on aerosol optical depth (AOD) evaluation in the “dust belt” region, a term that itself implies a strong emphasis on mineral dust aerosols due to the prevailing climatic and geographic conditions. While we acknowledge that other aerosol types may be present (e.g., sea salt, sulfate, biomass burning), the study area is globally recognized for frequent and intense dust activity, which naturally justifies the thematic focus.

Differentiating among aerosol types would require additional parameters—such as the Angstrom exponent or single scattering albedo—which are outside the scope of this study and not necessary for achieving the paper’s main objective: evaluating AOD retrieval performance from satellite and reanalysis datasets.

We also find the reviewer’s concern about a reader's interpretation somewhat misplaced. The title, abstract, and introduction clearly define the scope of the work, and any reader interested in broader aerosol classification would reasonably seek that in studies focused on aerosol type discrimination. Our manuscript does not claim to provide aerosol speciation but rather focuses on the quality of AOD data in a dust-dominant region.

Nonetheless, we are open to adding a clarifying sentence acknowledging the possible presence of other aerosol types and stating that this study concentrates on dust-dominant conditions as characteristic of the region.

 

Comment 10:  L. 81-82: Can the authors provide references of this claim?

Reply 10:         New references have been added

 

Comment 11:  L. 84: The sentence is a very misleading. Not all satellites are have polar orbits. Geosynchronous satellites exist and the data from these type of satellites are very important and can also observe dust in real time.

Reply 11:         We give respect to referee’s knowledge and believe in it. We also confirm that we have no such claim as pointed out by the learned referee. Nevertheless we have gone through the entire manuscript to ensure that no such sentence should be used.

 

Comment 12:  L. 88: The statement contradicts with the second statement in the paragraph.

Reply 12:         The alteration has been made.

 

Comment 13:  L. 99-101: Misleading statements. Ground instruments have limitations too. If there are clouds and dust event occur, sun photometer data are useless. Consider also the fact that sun photometers can't provide data at night even during dust events.

Reply 13:         We respectfully note that the language used in this comment is unnecessarily strong and appears to misrepresent our manuscript. At no point did we claim that ground-based instruments, such as AERONET sun photometers, are without limitations. On the contrary, our methodology explicitly uses Level 2.0 AERONET data, which are quality-assured and cloud-screened—precisely because we acknowledge that data collected during cloudy conditions or low solar angles (e.g., at night) are not usable for AOD analysis.

Our use of AERONET observations follows established scientific standards for satellite validation, and the inherent limitations of these instruments—such as lack of nighttime data and cloud interference—are well known in the remote sensing community. If needed, we can briefly restate these limitations in the revised manuscript to avoid any possible misinterpretation, though we do not believe any “misleading statements” exist as claimed.

 

Comment 14:  L. 145-147: Judging from the flow of thought (discussing dust to discussing aerosols), the authors have to state here that AOD of aerosols are not necessarily AOD from dust.

Reply 14:         We give respect to the learned referee’s point and have made changes in the revised manuscript. However, at the same time it is very basic thing for a researcher working in this field and should not require such statement. In addition, using words like “Judging from the flow of thought” seems to be inappropriate for such a minor thing.

 

Comment 15:  L. 157-162: Why is the effect of relative humidity in these observations not stated? How do relative humidity changes the AOD in each site. The authors must take into consideration that high AOD due to high RH does not mean high AOD from dust. Without this distinction, the discussion of the results presented in the manuscript is highly questionable!

Reply 15:         We acknowledge the reviewer’s concern regarding the influence of relative humidity (RH) on aerosol optical depth. However, we respectfully disagree with the assertion that the lack of a detailed RH-based analysis renders the discussion or findings questionable. While RH can indeed affect aerosol optical properties—particularly through hygroscopic growth in non-dust aerosols—addressing this variable in depth would require a separate analysis involving additional datasets, such as collocated in-situ or reanalysis humidity data, and potentially aerosol type classification, which is beyond the intended scope of this study.

Our manuscript clearly focuses on evaluating satellite and reanalysis AOD retrievals against AERONET observations in a region predominantly influenced by dust. It is neither feasible nor expected to incorporate every possible atmospheric variable into a single focused study, especially when doing so would deviate from our core objectives. We believe that pressing for such inclusion to the extent of questioning the validity of the entire analysis is not appropriate in the context of the study’s clearly defined goals.

That said, we do recognize the relevance of RH as a potential influencing factor on AOD and will consider briefly noting this in the revised discussion as a limitation and a point for future exploration.

 

Comment 16:  L. 233-236: Giovanni website is used only to visualize the data before serious analysis is done and NOT for downloading data. The Giovanni website is used to aid researchers not waste time working on large data set not knowing what results they should be expecting. After knowing the expected result (which is a rough picture presented in Giovanni website), data should be downloaded from respective archive, e.g., LAADS DAAC.

Reply 16:         We respectfully disagree with the reviewer’s rigid interpretation of the intended use of the Giovanni platform. Giovanni, developed and maintained by NASA GES DISC, is not only a data visualization tool but also a legitimate source for accessing preprocessed, gridded, and quality-controlled datasets that are widely used in peer-reviewed research. NASA itself supports and encourages the scientific use of Giovanni-based datasets, especially for regional-scale analyses, as in the case of our study.

Our choice to use data from Giovanni was based on its ability to provide spatially and temporally averaged MODIS AOD products in a consistent and efficient format. The datasets used were well-documented and fully referenced in the manuscript. While it is true that LAADS DAAC provides more granular Level 2 swath data, those are not necessary for the scope of this study, which focuses on broader spatial patterns and performance evaluation—not pixel-level retrieval diagnostics.

Thus, we assert that using Giovanni is scientifically valid and entirely appropriate for the nature of our analysis.

 

Comment 17:  L. 252-253: Better use the word "level" and indicate the level number so as to be consistent with how AERONET data are categorized.

Reply 17:         Changes have been made.

 

Comment 18:  L. 256: atmospheric clarity=> weird choice of word.

Reply 18:         We respectfully disagree with the reviewer’s characterization of the term "atmospheric clarity" as a “weird” choice. The phrase is commonly used in both scientific and general contexts to describe the degree to which the atmosphere is free from particulates, haze, or pollution—concepts directly relevant to aerosol optical depth. We believe it is an appropriate and intuitive expression in the context of discussing aerosol loading and visibility. However, if the editor recommends using a more specific or technical term, we are open to revising it.

 

Comment 19:  L. 259-260: What AERONET version and AOD level is used? The authors indicated that the AERONET data is retrieved from the AERONET website but the acknowledgment section contradicts this. Which is the true case?

Reply 19:         Already mentioned in the text.

 

Comment 20:  L. 290-291: Again what AERONET level is used?

Reply 20:         Already mentioned in the text.

 

 

Comment 21:  Figs. 2, 3 and 4: The authors did provided the state and end dates of the data presented in the graph, just the number of points (Table 1). Do the data have significant points for temporal analysis? Are there gaps (temporal) in the data that the reader should be aware of? What are the equations and r2 values of each graph?

Reply 21:         Values are already given in table 2.

 

comment 22:   L. 455 & 458: AERONET => Terra?

Reply 22:         Not sure what is the question.

 

Comment 23:  L. 470-471: The Angstrom coefficient from AERONET data should provide information on the presence of fine and coarse (dust) aerosols in the overall result rather than saying flatly "likey due to local pollution and complex...". This is missing in the analysis and the reader is just bombarded by generalities when specific information can be explicitly stated to increase scientific knowledge.

Reply 23:         We are fully aware of the utility of the Angstrom exponent in distinguishing between fine- and coarse-mode aerosols, and we have addressed this topic in detail in a separate publication, which was recently published in the journal Atmosphere. As experienced researchers in aerosol remote sensing, we made a deliberate decision to focus this manuscript on evaluating the performance of MODIS and MERRA-2 AOD retrievals against AERONET observations, rather than expanding into aerosol type classification, which would require a broader and more detailed dataset and discussion.

It is simply not feasible—nor expected—to include all possible dimensions of aerosol analysis in a single manuscript, especially when reviewers raise additional concerns over the number of authors involved. Questioning the capability of two researchers to carry out a comprehensive scientific study is not a constructive critique and detracts from a fair evaluation of the manuscript’s scientific merit.

 

Comment 24:  L. 492-494: Really? If this is the case, there should be some systematic error since the same algorithm is applied in retrieving the aerosol optical properties from the data from each site.

Reply 24:         We respectfully disagree with the implication that our result necessarily points to a systematic error in the retrieval algorithm. The reviewer’s concern seems to stem from the observation that “the same algorithm is applied in retrieving the aerosol optical properties from the data from each site.” However, our analysis is not merely a test of the retrieval algorithm itself in isolation, but rather an evaluation of the satellite-derived aerosol products against ground-based AERONET measurements over different geographical locations with distinct meteorological and aerosol regimes.

The bias differences we observed among Aqua, Terra, and MERRA-2 are influenced not only by the retrieval algorithm but also by the local aerosol type, surface reflectance characteristics, and atmospheric conditions at each site. As is well known, the performance of aerosol optical depth (AOD) retrievals can vary depending on factors such as surface brightness, relative humidity, and aerosol composition — all of which vary geographically and seasonally. Therefore, the observed differences in bias magnitude are not necessarily indicative of a systematic error inherent to the retrieval algorithm, but rather reflect how the algorithm interacts with the local conditions.

Furthermore, the statistical analyses applied in our manuscript explicitly account for these regional differences by assessing multiple sites with diverse characteristics. Our findings of consistent RMSE and MAE across platforms actually support the relative stability of the retrieval algorithm, despite varying biases that can be attributed to environmental factors.

 

Comment 25:  L. 505-508: Should these results verify the results in Sec. 4.1?

Reply 25:         We thank the reviewer for raising this question. However, it is not entirely clear to us what specific concern the reviewer intends to highlight here. Section 4.1 of the manuscript presents the validation of the satellite-derived aerosol optical properties against AERONET measurements, providing an overall assessment of accuracy and error metrics. In contrast, the current section focuses on the spatial analysis of these properties across different geographical regions, which builds on the validation findings but is not designed as an independent verification of Section 4.1.

In other words, the spatial analysis is meant to explore the geographical patterns and regional differences in retrieval performance, taking into account the site-specific biases and errors already characterized in the validation section. Therefore, the results presented here are complementary to, and consistent with, the findings of Section 4.1, but they are not intended as a separate verification of those results.

 

Comment 26:  L. 509-514 (and also for other similar paragraphs describing the datasets from other sites): The information of the dataset's time, month , year is not stated. The authors should note that the data presented here includes the effects of relative humidity which increases AOD but does not mean "more aerosols".

Reply 26:         We appreciate the reviewer’s observation that relative humidity (RH) influences aerosol optical depth (AOD) by increasing particle size through hygroscopic growth, which does not necessarily correspond to an actual increase in aerosol mass or number. This is a valid point and an important aspect of interpreting AOD measurements.

However, we would like to clarify that the primary focus of the present study is the performance evaluation of satellite-retrieved AOD products against AERONET observations and their spatial patterns, rather than a detailed investigation of the meteorological drivers of AOD variability, such as RH. While RH undoubtedly plays a role in the measured and retrieved AOD values, explicitly isolating and quantifying its contribution was beyond the scope of this work.

 

Comment 27:  L. 531-533: So what are in these sites that produce different results compared to other sites? Is it due to more aerosol sources, aerosol growth or dust (which can be confirmed from AERONET data)?

Reply 27:         We thank the reviewer for this insightful question regarding the factors contributing to the site-to-site differences in results. As the reviewer correctly suggests, the observed differences are indeed influenced by a combination of local aerosol sources, aerosol growth processes, and the presence of dust — all of which vary geographically and can be inferred from AERONET data.

In our analysis, we did not explicitly separate the contributions of these factors, as the primary objective was to evaluate the overall performance of the satellite products across diverse regions. However, we acknowledge that local characteristics such as urban/industrial emissions, biomass burning, desert dust, and meteorological conditions (including relative humidity and wind patterns) can all affect both the magnitude and type of aerosols present, and therefore the retrieval performance.

 

Comment 28:  L. 533-534: This is a misleading/ambiguous statement. What statistical parameters do the authors refer to?

Reply 28:         We thank the reviewer for pointing this out. We understand the concern that the original wording may come across as ambiguous and could benefit from greater clarity. Our intent was to convey that the spatial analysis findings align with the statistical error metrics (bias, RMSE, MAE) reported in the validation section, in the sense that regions/sites with higher or lower performance in the statistical validation also exhibited corresponding spatial patterns.

We also note that while the language in the reviewer’s comment was strongly worded, we appreciate the underlying point and have clarified the manuscript accordingly.

 

Comment 29:  Figs. 5, 6 and 7: The colorbar is missing. Are the colors in the each figure have the same minimum and maximum values? If not, then the discussion has to be re-evaluated. Which of the graphs are a), b), c), etc.? What are the years that represent these data? What are the standard deviation graphs?

Reply 29:         These points are already present in the manuscript.

 

Comment 30:  L. 681-682: Theoretically, what values of optical properties (e.g., AOD) are to be expected under these type of environments? I think the reader has to know this information so he can evaluate the theoretical and the results presented in this paper.

Reply 30:         We thank the reviewer for raising this point. However, we respectfully note that providing “theoretical” or assumed values of optical properties such as AOD for the studied environments is not straightforward or particularly meaningful in the context of this study. This is because AOD is a highly variable quantity that depends on real-time local emissions, meteorology, and aerosol type, all of which can fluctuate significantly even within the same site and season.

Instead of relying on theoretical estimates, our study leverages the availability of AERONET ground-based observations at each site, which provide high-quality, site-specific measurements of AOD. These serve as the reference against which the satellite products are evaluated. We believe this approach is more appropriate and robust than speculating on general “expected” values, which may not reflect actual local conditions during the study period.

 

Comment 31:  L. 692-699: A short description of each site may help the reader understand the results.

Reply 31:         We are sorry but this may lead to increase the length of manuscript.

 

Comment 32:  L. 694-695: References to this claim?

Reply 32:         New references are already added

 

Comment 33:  L. 713: References to this claim?

Reply 33:         New references are already added

 

Comment 34:  Figs. 8, 9, 10: Whey is MERRA AOD higher than AERONET? Are the mean values within the standard deviation of each measurement point? Why is the standard deviation not shown? Why display annual result when the authors' objective is to know how MODIS and MERRA measurements differ from AERONET measurements.  Why not compare daily dataset and study the difference of these daily differences from these measurements?

Reply 34:         We thank the reviewer for these thoughtful and detailed questions regarding the comparison of MERRA and AERONET AOD, the role of standard deviation, and the use of annual versus daily datasets.

We acknowledge that investigating why MERRA AOD tends to be higher than AERONET at certain sites, whether the mean differences fall within the standard deviation of the measurements, and how the daily differences behave statistically are all interesting and valuable research questions. However, these analyses involve a deeper exploration of the uncertainties, error propagation, and temporal dynamics of the datasets, which go beyond the scope of the current study.

The primary objective of this paper is to evaluate the overall performance and spatial patterns of MODIS and MERRA AOD relative to AERONET, using annual averages and aggregate error metrics. This approach provides a broad assessment of retrieval performance and regional variability, as stated in the manuscript. A thorough daily-level analysis, with a focus on standard deviations and variability of differences, would indeed be a meaningful next step and merits a dedicated, separate study to do it justice.

 

Comment 35:  Figs. 11, 12 and 13: The wavelength information of the AOD is not stated. Which wavelength is used? The results in the graphs seem to contradict the results in the previous figures (Figs. 8, 9 and 10) that show that AOD from MERRA is greater than the AOD from AERONET.

Reply 35:         We respectfully disagree with the reviewer’s implication here. The wavelength information for the AOD used in Figs. 11, 12, and 13 is indeed clearly stated in the figure captions and the methodology section — the comparisons are all performed at 550 nm, consistent throughout the manuscript. We suspect that the reviewer may have inadvertently overlooked this information when raising this point.

Regarding the perceived contradiction between Figs. 8–10 and Figs. 11–13: the figures show different aspects of the data. Figs. 8–10 present aggregated annual averages and bias metrics, while Figs. 11–13 illustrate the spatial distribution and site-level comparisons. Both sets of figures are fully consistent with each other, reflecting the same trend that MERRA AOD tends to be higher than AERONET in most regions.

We have carefully reviewed the figures and the text again and confirm that there is no inconsistency in the results presented. To avoid potential confusion for readers, however, we have revised the figure captions to explicitly state the wavelength and added a clarifying sentence in the results section to emphasize the consistency between these sets of figures.

We trust this clarifies the reviewer’s concern.

 

Comment 36:  L. 979 (and all related statements describing the AOD at each site): What are the standard deviation values? Are the mean AOD values at each site presented here show significant difference from the mean AOD values from other sites? Why?

Reply 36:         We thank the reviewer for raising this point. However, we respectfully note that the standard deviation (SD) values are already clearly presented in Table 2 of the manuscript, as part of the summary statistics for each site. We believe this adequately addresses the question about the variability of AOD at each site.

Regarding the second part of the comment — why mean AOD values differ across sites — this reflects the well-known fact that aerosol loading is highly site-specific, influenced by local and regional emission sources, land cover, meteorology, and aerosol type. It is entirely expected and consistent with the literature that each site exhibits distinct mean AOD values, and this is indeed one of the motivations for conducting spatially resolved validation studies.

 

Comment 37:  L. 1003-1004 (and all related statements describing the years with high and low AOD): Why are the years of maximum and minimum AOD values different? When instruments gather data from the same parameter, the results should show the same trend of high and low values. There maybe differences but the trend should be the same.

Reply 37:         We thank the reviewer for this observation. We agree in principle that instruments measuring the same parameter over the same period should generally capture the same overall trends in variability (e.g., seasonal or interannual peaks and lows). However, it is also well documented in the literature that differences in retrieval algorithms, spatial representativeness, temporal coverage, and sensitivity to surface and atmospheric conditions can lead to discrepancies in the specific years when maximum or minimum AOD values are recorded by different instruments.

For example, satellite-based and reanalysis products such as MODIS and MERRA have different spatial footprints, observation times, and assumptions about aerosol properties compared to AERONET. These differences can influence which specific year exhibits the extreme values in each dataset, particularly when the differences between years are small and within the measurement uncertainties.

 

Comment 38:  L. 1008: Indicate/define the "spring months" here.

Reply 38:         Done

 

Comment 39:  L. 1015-1017 (all related statements describing the AOD values from AERONET and MERRA and MODIS): The positioning of the graphs is does not follow the narrative. The authors may need to rearrange the graphs to fit the narrative in the text.

Reply 39:         Done

 

Comment 40:  L. 1027: How hot? The authors need to have specific description.

Reply 40:         This is a meteorological term and we are not measuring the temperature of a specific object

 

Comment 41:  L. 1038-1040: References (ground-based measurements) to support this statement is needed.

Reply 41:         Added

 

Comment 42:  L. 1058-1059: Quantitative values of temperature and humidity are not stated. How are these temperatures compare to other sites. The reader does not have an idea of how these temperature and humidity values differ from other sites.

Reply 42:         We thank the reviewer for raising this point. We agree that temperature and humidity can influence aerosol properties through processes such as hygroscopic growth and dispersion. However, the primary focus of this study is on the evaluation of satellite and reanalysis AOD retrievals relative to AERONET, rather than a detailed meteorological intercomparison across sites.

Quantitative values of temperature and humidity for each site are therefore not explicitly presented, as they are not central to the scope of this work. Instead, the variability of AOD across sites indirectly reflects the influence of local meteorological conditions along with emission sources and aerosol types. A detailed cross-site meteorological comparison would be a valuable topic for a separate, dedicated study.

 

Comment 43:  L. 1060-1061: The authors state the same reasons in other sites. Which site then does not have secondary aerosol formation under the same season? Why?

Reply 43:         We thank the reviewer for this observation. Our intention in noting “secondary aerosol formation” at multiple sites was to highlight that this process is a common feature influencing AOD in many environments, particularly during seasons with high photochemical activity and elevated humidity. We did not mean to imply that every site experiences this phenomenon to the same extent or under identical conditions.

It is well understood that the intensity of secondary aerosol formation varies with local precursor emissions (e.g., SO₂, NOₓ, VOCs), meteorological conditions, and regional transport. Some sites, especially those with lower anthropogenic emissions or dominated by coarse-mode natural aerosols such as desert dust, may exhibit comparatively less secondary aerosol formation during the same season. However, disentangling and quantifying the relative contributions of secondary formation versus other sources was not the primary objective of this study and requires a more detailed chemical composition analysis beyond the scope of this paper.

 

Comment 44:  L. 1076: What geographical characteristics? The authors need to be specific in this description. If the authors collaborated with the PI of each site, they would have a better knowledge of the site's climate/weather characteristics and may be weather data to support the authors' claims in the text.

Reply 44:         We respectfully disagree with the reviewer’s implication regarding the credibility of our work. The geographical and meteorological characteristics referenced in the manuscript are not speculative, nor do they depend on personal collaboration with site PIs. Rather, they are based on well-documented information available in the literature and in publicly accessible datasets describing the environments where the AERONET sites are located.

Our intention in mentioning “geographical characteristics” was to acknowledge that site-specific factors such as land cover (urban, desert, coastal), dominant aerosol type (dust, biomass burning, anthropogenic pollution), and local meteorology (humidity, wind patterns) play an important role in shaping aerosol properties and therefore influence the agreement between satellite/reanalysis products and AERONET. These factors are widely recognized in aerosol studies and do not require site-specific collaboration to establish.

 

 

Comment 45:  L. 1077:  "prone to dust activity..." => So does other sites as mentioned in earlier paragraphs. Do the AERONET data support this claim?

Reply 45:         We thank the reviewer for this observation. The statement regarding sites being “prone to dust activity” is supported by the geographical setting of these locations, which are well-documented in the literature as being influenced by regional dust transport. This is also consistent with the AERONET records, which often report higher coarse-mode fractions and elevated AOD episodes corresponding to dust events.

 

 

Comment 46:  L. 1080: "local farming practices..." => Not existing on other sites? How much of this activity is present in each site? Are there references to support the authors' claim?

Reply 46:         We thank the reviewer for the question. The reference to “local farming practices” was made to highlight a site-specific factor influencing seasonal aerosol emissions at certain locations. We do not claim that such practices exist uniformly across all sites, but rather that they are more relevant in agricultural regions where biomass burning or soil disturbance contributes to aerosol loading.

 

Comment 47:  L. 1084-1085: "regional influences" => references?

Reply 47:         Added

 

Comment 48:  L. 1100: "dust activity" => This can be verified from the AERONET data. Does the AERONET data support this claim by the authors?

Reply 48:         We thank the reviewer for this observation. Yes, the claim regarding dust activity is supported by AERONET data. The AERONET inversion products for the relevant sites frequently indicate elevated coarse-mode aerosol fractions and high AOD episodes, which are consistent with dust events reported in the literature.

 

Comment 49:  L. 1104-1105: These descriptions (and also of other sites) are too general. The authors should provide references to support these descriptions and claims.

Reply 49:         Added

 

Comment 50:  L. 1112: "year 2015" => Why 2015? What's the cause?

Reply 50:         Added

 

Comment 51:  L. 1114-1117: The authors need to provide references to support this claim.

Reply 51:         Added

 

Comment 52:  L. 1123-1126: These descriptions are too general and does not directly give scientific proof.

Reply 52:         We thank the reviewer for this remark. We respectfully note, however, that the purpose of the descriptive statements in this section was to provide contextual interpretation of the results, linking observed AOD patterns with well-documented geographical and meteorological factors. These are not intended to serve as stand-alone “proof,” but rather as scientifically reasonable explanations supported by established knowledge and the AERONET observations used in this study.

 

Comment 53:  Conclusion: The conclusion can be improved after major revision of the paper. The authors should directly point out if the objectives are achieved and respond to the the question "Are the results scientifically significant?"

Reply 53:         We thank the reviewer for this suggestion. We agree that the conclusion section can be strengthened by more explicitly addressing whether the study objectives have been achieved and by highlighting the scientific significance of the results.

Accordingly, we have revised the conclusion to clearly state that the study objectives — namely, the evaluation of MODIS and MERRA AOD retrievals against AERONET across diverse geographical regions, and the analysis of their spatial variability — have been achieved. We also emphasize the scientific significance of our findings in demonstrating the relative stability of the retrieval algorithms across platforms, while also identifying the influence of regional aerosol and meteorological conditions on performance.

 

Comment 54:  The quality of the usage of the English language can be improved. The authors should not use misleading and ambiguous words or sentences.

Reply 54:         We respectfully note the reviewer’s comment regarding the quality of the English language. As the corresponding author of more than 100 peer-reviewed research articles published in well-established international journals, with over 1000 citations, we are confident in our ability to communicate scientific findings clearly and professionally.

Nevertheless, we recognize the importance of ensuring that the manuscript reads as clearly as possible for a broad audience. To this end, we have carefully reviewed the entire text to remove any potentially ambiguous or unclear wording, and we have revised several sentences for improved clarity and precision.

We trust that these refinements adequately address the reviewer’s concern while maintaining the scientific rigor and readability of the manuscript.

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The necessary corrections have been made and the current version is much improved over the previous one.

Author Response

We really appreciate the useful comments made by the learned referee and the confidence in our manuscript. 

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript presents the performance of MERRA-2 and MODIS in AOD inversion for dust-affected areas. The revised manuscript already has a good level of publication quality. I only have a few minor comments.

1）The abbreviations used in this article are not standardized. In general, the R² represents the determination coefficient, and the R represents the correlation coefficient.

2) The abstract states that the AOD performance of MERRA-2 is unexpectedly superior to that of MODIS, and this has not been discovered or reported in other studies, such as 10.3390/rs14040821, 10.5194/acp-23-8187-2023, and 10.1016/j.rse.2021.112410. Have you considered the differences in resolution between these two products, and have you conducted appropriate quality control for the satellite products?

3) I noticed that the author seemed to have conducted only research on the region of dust environment there. This point must be clearly indicated in the title and the abstract.

4) Please clarify how the IOA is defined and calculated.

5) Based on AERONET data, although the R² value of MERRA-2 is slightly higher than that of MODIS, MERRA-2 also shows a greater negative deviation. This point should be emphasized.

Author Response

Comment 1:    The abbreviations used in this article are not standardized. In general, the R² represents the determination coefficient, and the R represents the correlation coefficient.

Reply 1:                       The changes have been made in the revised manuscript.

 

Comment 2:    The abstract states that the AOD performance of MERRA-2 is unexpectedly superior to that of MODIS, and this has not been discovered or reported in other studies, such as 10.3390/rs14040821, 10.5194/acp-23-8187-2023, and 10.1016/j.rse.2021.112410. Have you considered the differences in resolution between these two products, and have you conducted appropriate quality control for the satellite products?

Reply 2:                       Yes indeed we have conducted quality control of satellite products. However, this performance is associated with temporal resolution which has been discussed in the revised manuscript.

 

Comment 3:    I noticed that the author seemed to have conducted only research on the region of dust environment there. This point must be clearly indicated in the title and the abstract.

Reply 3:                       As suggested by the learned referee the changes have been made in the revised manuscript.

 

Comment 4:       Please clarify how the IOA is defined and calculated.

Reply 4:  It has been clarified in the revised manuscript.

 

Comment 5:    Based on AERONET data, although the R² value of MERRA-2 is slightly higher than that of MODIS, MERRA-2 also shows a greater negative deviation. This point should be emphasized.

Reply 5: This point is included in the conclusion section of the manuscript.

 

Reviewer 3 Report

Comments and Suggestions for Authors

Please see attached file.

Comments for author File: Comments.pdf

Author Response

The file has been attached

Author Response File: Author Response.pdf