Review Reports - Spatial–Temporal Variation and the Influencing Factors of NO2 Column Concentration in the Plateau Mountains of Southwest China

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

1- Please check the instruction of this jouurnal (e.g., number of words in the Abstract section.).

2- What do you mean by "Geographical detector" in the Keeyword section?

3- Add quantitative results to the Abstract and Conclusion section.

4- What is the main innovations of this paper. Please highlight them in the Abstract and Introduction section. I am not convinced from innovation of this paper.

5- Add a sub-figure to figure 1 to show where your study area located on a global sphere.

6- Add a figure to show the general workflow of your methodology.

7- Change the structure of Conlcusion section to fit better with other published paper in this journal.

8- What is the future direction and challenges faced in this paper?

9- The quality of the figures and texts in them must be revised.

10- In Figure 6, what is the vertical trend in subfigures b-d?

Author Response

Greetings! I am truly honored to have received the review of my manuscript from a seasoned scholar like you. Your profound insights into the field of research and meticulous attention to detail are immensely encouraging and motivating for me. Each suggestion you have offered is of great value, not only targeted but also uniquely insightful, playing an indispensable role in enhancing the academic standards and practical significance of my paper. Inspired by your feedback, I have thoroughly revised the manuscript.Moreover, your critique regarding potential logical fallacies and unclear expressions in the paper has been immensely beneficial. I have carefully corrected these parts and provided clear explanations to ensure the accuracy and readability of the paper. I appreciate your selfless dedication and valuable time.

I look forward to receiving more of your opinions and suggestions. Thank you wholeheartedly for your assistance and support.Here is our response to the review comments

Please check the instruction of this jouurnal (e.g., number of words in the Abstract section.).

Thank you for pointing this out. We agree with this comment and have made the necessary modifications.

What do you mean by "Geographical detector" in the Keeyword section?

L37:Thank you for pointing this out, we agree with this comment. We have changed 'Geographic detector' to 'Geographic detector model'

Add quantitative results to the Abstract and Conclusion section.

Thank you for pointing this out. We agree with this comment and have made the necessary modifications. Revise the abstract and conclusion sections and add quantitative results.

What is the main innovations of this paper. Please highlight them in the Abstract and Introduction section. I am not convinced from innovation of this paper.

Thank you for pointing this out; we agree with this comment. We explain that the innovation of this paper lies in the following aspects:

Given the complex terrain and economic development conditions in Guizhou Province, research on the tropospheric NO₂column concentrations using satellite remote sensing is still insufficient. Therefore, observing the temporal and spatial evolution characteristics of tropospheric NO₂column concentrations can ensure the stable development of atmospheric environmental quality.

(2) The temporal and spatial distribution of tropospheric NO₂column concentrations is influenced by a combination of factors, including human activities, topography, and climate. Currently, most studies on plateau and mountainous areas focus solely on the impact of socio-economic or natural factors, with relatively few studies conducting comprehensive analyses that integrate both aspects. This paper utilizes geographic detectors to explain the extent and nonlinear comprehensive effects of natural and social factors on the temporal and spatial distribution of NO₂column concentrations.

Add a sub-figure to figure 1 to show where your study area located on a global sphere.

Figure 1:Thank you for pointing this out. We agree with this comment and have made the necessary modifications.

Add a figure to show the general workflow of your methodology.

Figure 2: Thank you for pointing this out. We agree with this comment and have made the necessary modifications.

7- Change the structure of Conlcusion section to fit better with other published paper in this journal.

Thank you for pointing this out. We agree with this comment and have made the necessary modifications

8- What is the future direction and challenges faced in this paper?

Section 5.1.4: Thank you for pointing out this point. We agree with this comment and have made revisions to Section 5.1.4. We propose that the future direction of this article is to consider using multi-source data such as OMI satellite data, GEMS satellite data, and ground NO2 monitoring data for fusion analysis in future research, in order to improve the accuracy of tropospheric NO2 column concentration research,

The quality of the figures and texts in them must be revised.

Thank you for pointing this out. We agree with this comment and have made the necessary modifications.

In Figure 6, what is the vertical trend in subfigures b-d?

Figure 6: Thank you for pointing this out. We agree with this comment and have made modifications. We explain that the Moran's I values mentioned in Figure 6 (e.g., 0.99, 0.81, 0.70) represent the global Moran's I index, which measures the spatial autocorrelation of the entire spatial dataset. High values of the Moran's I index indicate strong positive spatial autocorrelation. The scatter points in Figure 6 represent the distribution pattern of the local Moran's I index, which measures the spatial autocorrelation between each individual observation and its neighboring values. The vertical trend is due to local patterns in certain areas; the upper part of the scatter plot in Figure 6 (high values) shows a clustering state, indicating that areas with high NO₂concentrations tend to be adjacent to other high-concentration areas. In contrast, the lower part of the scatter plot for the years 2020-2022 (low values) exhibits a vertical distribution trend, reflecting a weakened aggregation effect of low NO₂ column concentration areas. Therefore, to obtain insights into the spatial variations of tropospheric NO₂ column concentration within Guizhou Province, we conducted research using cold and hot spot spatial analysis and local spatial autocorrelation statistics. This pattern may be related to geographic location, environmental factors, or human activities. The description information for Figure 6 is as follows:

1.X-axis: Represents the standardized values of the observations, i.e., the original observation values minus their mean, divided by the standard deviation.

2.Y-axis: Represents the standardized values of the spatial lag of the observation, i.e., the weighted average of the neighboring values of the observation, minus its mean, divided by the standard deviation.

The scatter plot of the local Moran's I indices can reveal the following information:

3.Spatial clustering: If the scatter points are concentrated in the upper right or lower left corners of the plot, it indicates positive spatial autocorrelation, meaning similar values tend to cluster together.

4.空间异常值：如果散点位于左上角或右下角，则表示负空间自相关，这意味着相似的值往往彼此分散。

5.随机分布：如果散点在图中随机分布，没有任何显著的聚类或离散模式，则说明空间数据没有表现出显著的空间自相关。

6.趋势：如果散点显示某种趋势（如垂直或水平），这可能表明数据中存在一些系统性的空间模式。

Author Response File: Author Response.pdf

Reviewer 2 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

Please check the attached file.

Comments for author File: Comments.pdf

Author Response

I would like to express my heartfelt gratitude for taking the time out of your busy schedule to review my paper. Your professional insights and constructive suggestions have been crucial in enhancing the quality of my research. I have carefully considered each of your points and made diligent revisions and improvements to the paper accordingly.Your profound understanding of research methodology and detailed advice on data analysis have allowed me to view my work from new perspectives, leading to necessary improvements in certain key areas. Additionally, your valuable feedback on the structure and logical flow of the paper has greatly assisted me in enhancing the overall clarity and coherence of my writing.

I am particularly grateful for your guidance on potential shortcomings in the paper, which not only benefits my current research but also provides valuable reference for my future work. I will continue to strive to ensure that my research outcomes are more rigorous and reliable.

Once again, thank you for your hard work and selfless dedication. Please continue to oversee my research progress, and I look forward to more opportunities for communication and collaboration in the future. We sincerely wish you all the best in your life and work.

Section 3.1

I'm afraid I still have to disagree with this comparison's results and deem it to be flawed for

the following reasons:

Comparison between two quantities with different units is not correct.

Thank you for pointing this out. We have made modifications. Our explanation is that existing studies have shown a strong correlation between tropospheric NO₂ column concentration and ground-level NO₂ concentration. Research by Judd et al. [1] indicates that although the TROPOMI sensor underestimates NO₂ column concentrations by 7-19%, the consistency between its retrieval results and ground monitoring values reaches 85% on a global scale. Additionally, numerous scholars have used linear fitting to validate the applicability of NO₂ column concentration data against ground data, achieving satisfactory results [2-5]. Therefore, this paper uses the ground-level NO₂ concentrations from air quality reports of nine cities in Guizhou Province as ground monitoring results and conducts correlation analysis and linear fitting with the NO₂ column concentrations.

(2) There is no description of the ground-based measurements. What is the sites' topography

(i.e., urban, rural, suburban)? How was the dataset filtered and analyzed? Continuous

monitoring data is associated with missing values. How did the authors deal with this issue?

How many monitoring stations were used?

L178-180：Thank you for pointing this out. We agree with this comment and have made the necessary modifications. We explain that this article collects monthly average NO₂ concentration data from urban ground monitoring in 9 cities and prefectures in Guizhou Province from 2019 to 2022 through the China Air Quality Monitoring Platform. The data is continuous monitoring monthly average data with 33 monitoring stations.

(3) The authors used satellite concentration data, which means they have utilized the profiles

for analysis. Are satellite data representative of the surface(i.e., ~100 meters)? Or has the

data from a certain height (i.e., 1km, 2km, 3km, etc.) been selected? Or does the data

represent the whole troposphere?

Thank you for pointing this out. We agree with this comment and explain that the Sentinel-5P NO₂ product used in this article represents the concentration data of the entire troposphere.

What does inverted value/data mean?

Thank you for pointing this out. We agree with this comment, but it is not what we intended to express. We have removed the phrase 'inverted value/data'.

(5) There is no discussion on the biases that could affect the comparison result. The results

clearly show a strong bias towards the satellite observations.

Thank you for pointing this out. We agree with this comment and have made modifications. We will explain it from three aspects:

(1) The relationship between ground monitoring NO₂ data (x-axis) and remote sensing monitoring NO₂ data (y-axis). The equation of the linear fitting line is typically represented as Y = mx + b, where m is the slope and b is the intercept. In this case, the equation Y = 0.72 + 0.15x indicates that the remote sensing monitoring NO₂ data (Y) is a linear function of the ground monitoring data (x), with a slope of 0.15 and an intercept of 0.72.

The R² value, also known as the coefficient of determination, measures the goodness of fit of the linear model. An R² value of 0.752 means that 72.5% of the variability can be explained by the linear relationship, which indicates a relatively strong fit.

Therefore, the significance of the red line is as follows:

It illustrates the best linear relationship between ground monitoring NO2 data and remote sensing monitoring NO₂ data.

It helps visualize the correlation between the two data sets.

It provides a quantitative metric for assessing the consistency between ground monitoring and remote sensing data.

The range of the remote sensing NO₂ column concentration data values is primarily concentrated between 3 and 3.75, while the range of ground monitoring NO₂ values is mainly concentrated between 15 and 20. After adjusting the scale of the x-axis, the trend line leans towards the ground monitoring data. Thus, we interpret that the fitting results of the two data sets do not favor the satellite remote sensing data. The figure below shows the results obtained after adjusting the scale.

(2) Based on the results from Image 3 in the manuscript, we can analyze from the following aspects:
(a) Slope of the linear fit:
If the slope of the linear fit is close to 1, it indicates a strong consistency between the remote sensing monitoring data and ground monitoring data, with no significant bias. If the slope is much less than 1, this may indicate that the remote sensing monitoring data is lower than the ground monitoring data, showing a negative bias in satellite observations. Conversely, if the slope is much greater than 1, this may suggest that the remote sensing monitoring data is higher than the ground monitoring data, indicating a positive bias in satellite observations.
(b) Intercept of the linear fit:
The intercept should not be zero; if it is significantly different from zero, this may indicate a systematic bias in the remote sensing monitoring data.
(c) R²:
The R² value provides an indicator of goodness of fit. The closer the value is to 1, the better the fit, indicating a higher consistency between the remote sensing data and ground data. A lower R² value may suggest that there are significant differences between the datasets, or that the remote sensing data cannot fully capture the variations in the ground monitoring data.
(d) Scatter distribution:
Observe the distribution of data points on the scatter plot. If the data points are closely clustered around the fitted line, it indicates a good consistency between the remote sensing data and ground data.
(e) Significance test:
Linear regression analysis is typically accompanied by significance tests. If the linear relationship is significant, it increases the credibility of the relationship between the remote sensing data and ground data.
(3) We compared and referenced the research results of other scholars:
(a) Zheng Zihao et al. in "Analysis of spatiotemporal changes of NO₂ pollutants in the Guangdong-Hong Kong-Macao Greater Bay Area based on Sentinel-5P" compared results.

(b) The results of Zheng Zihao's analysis of the spatiotemporal changes of NO₂ pollutants in the Guangdong Hong Kong Macao Greater Bay Area based on Sentinel-5P:

(c) The comparative results of Yu Ling's analysis of the spatiotemporal variation patterns and influencing factors of NO₂ in Guangxi show that:

(6) 'Evenly distribution on either side of the regression line."- I disagree with this statement.

The regression line is strongly biased toward the satellite observations.

Thank you for pointing out this point. We agree with this comment and have removed the unclear statement about "even distribution on both sides of the regression line".

(7) I presume the choice of the –(a)profile height and (b) the elevation of the study region

could be the potential reason for the overestimation in the satellite datasets. Theoretically, it

should be lower than surface observation (also seen in many observational studies). No

discussion on this critical issue has been made.

Thank you for pointing out this point. We agree with this comment. We explain that there may be some discrepancies between satellite transit time and local actual observation time, as well as some deviations between remote sensing monitoring data and ground monitoring data due to meteorological conditions such as cloud cover and aerosols. However, in this article, the fitting coefficient R² of the two sets of data is 0.752, which has good linear correlation and consistency. Many scholars have also confirmed the rationality and reliability of using tropospheric NO₂ column concentration instead of near ground large NO₂ concentration [4-5]. Therefore, this article uses the atmospheric NO₂ column concentration data inverted by Sentinel-5p satellite for subsequent analysis.

Section 3.4

There is no description of how to interpret Moragn's index

Thank you for pointing this out. We agree with this comment and have made modifications. We explain that the Moran's I values mentioned in Figure 6 (e.g., 0.99, 0.81, 0.70) represent the global Moran's I index, which measures the spatial autocorrelation of the entire spatial dataset. High values of the Moran's I index indicate strong positive spatial autocorrelation. The scatter points in Figure 6 represent the distribution pattern of the local Moran's I index, which measures the spatial autocorrelation between each individual observation and its neighboring values. The vertical trend is due to local patterns in certain areas; the upper part of the scatter plot in Figure 6 (high values) shows a clustering state, indicating that areas with high NO₂concentrations tend to be adjacent to other high-concentration areas. In contrast, the lower part of the scatter plot for the years 2020-2022 (low values) exhibits a vertical distribution trend, reflecting a weakened aggregation effect of low NO₂ column concentration areas. Therefore, to obtain insights into the spatial variations of tropospheric NO₂ column concentration within Guizhou Province, we conducted research using cold and hot spot spatial analysis and local spatial autocorrelation statistics. This pattern may be related to geographic location, environmental factors, or human activities. The description information for Figure 6 is as follows:

1.X-axis: Represents the standardized values of the observations, i.e., the original observation values minus their mean, divided by the standard deviation.

The scatter plot of the local Moran's I indices can reveal the following information:

4.Spatial outliers: If the scatter points are located in the upper left or lower right corners, it indicates negative spatial autocorrelation, meaning similar values tend to be dispersed from each other.

5.Random distribution: If the scatter points are randomly distributed in the plot without any significant clustering or dispersion patterns, it indicates that the spatial data does not exhibit significant spatial autocorrelation.

6.Trends: If the scatter points show some trend (such as vertical or horizontal), this may indicate the existence of some systematic spatial patterns in the data.

(2) L390: As a researcher, I believe this translation adheres to the grammatical and syntactical

standards appropriate for English academic papers. What does this mean?

L390: Thank you for pointing out this point. We have revised the content of the manuscript to "Using spatial cold and hot spot analysis and local spatial autocorrelation statistical methods, the study investigates the spatial variation of tropospheric NO₂ column”

(3) I cannot conceive the differences in the analysis in sections 3.2 and 3.3. The spatial

characteristics have already been established in section 3.2; thus, I find no valid reason for

the analyses in section 3.3

3.2.2 Thank you for pointing out this point. We have deleted the content on the annual variation of NO₂ column concentration in each city and state in section 3.2.2, revised the spatial distribution of NO₂ column concentration to section 3.2.2, and made modifications and improvements to the content of this section.

Section 5.1

This whole section is the repetition of sections 3 and 4 and is completely redundant. I find

no reason/significance in this section.

Thank you for pointing this out. We have revised Section 5.1 and provided explanations:

5.1.1: Our aim is to express the spatiotemporal distribution characteristics of NO₂ column concentrations in the highland mountainous region of Guizhou Province. The spatial distribution pattern of NO₂ column concentrations is consistent with the spatial characteristics of economic development. By conducting statistical analyses in conjunction with socioeconomic data, we draw conclusions. Regarding the analysis of cold and hot spots and the clustering results of high and low values, we use Guiyang City and Qiandongnan Prefecture as examples to illustrate the impact of human activities on local NO₂ column concentrations. We propose scientific monitoring and prevention measures based on the hot and cold regions of NO₂ concentrations and areas of high value aggregation. In terms of the temporal variation characteristics, we explain that on a seasonal scale, NO₂ column concentrations are higher in winter and lower in summer. Referring to previous studies , we suggest that the seasonal variations in NO₂ column concentrations in Guizhou Province are mainly due to differences in meteorological conditions, and we elaborate on this. Based on the changes in our annual average values, we explain that overall, from 2019 to 2022, the annual average NO₂ column concentrations in Guizhou Province remained at a low level and were relatively stable. Compared to the air pollution issues caused by urbanization and industrialization in mid-eastern cities of China, Guizhou has a relatively low population density and limited industrial development.

5.1.2: We have modified Section 5.1.2 to further explain the effects of meteorological factors and socioeconomic activities on NO₂ column concentrations based on the results of geographic detectors.

Section 5.1.3:

No idea why this is section is required?

Thank you for pointing out this point. We have made revisions to section 5.1.3.We explain that the NO₂ column concentration in the troposphere of Guizhou Province is closely related to the socio-economic development patterns of various regions, and that there are differences in socio-economic development between cities and states. The factors affecting NO₂ column concentration are a complex issue in different administrative units. Therefore, we take Zunyi City as an example for analysis and provide a certain degree of explanation and expansion for future research and analysis.

reference:

Judd, L.M.; Al-Saadi, J.A.; Szykman, J.J.; Valin, L.C.; Janz, S.J.; Kowalewski, M.G.; Eskes, H.J.; Veefkind, J.P.; Cede, A.; Mueller, M. Evaluating Sentinel-5P TROPOMI troposphericNO₂ column densities with airborne and Pandora spectrometers near New York City and Long Island Sound. Atmospheric Measurement Techniques 2020.
Zihao, Z.; Zhifeng, W.; Yingbiao, C.; Zhiwei, Y. Analysis of spatiotemporal changes ofNO₂ pollutants in the Guangdong Hong Kong Macao Greater Bay Area based on Sentinel-5P. Chinese Environmental Science 2021, 41, 63-72.
Wang, C.; Wang, T.; Wang, P.; Wang, W. Assessment of the performance of TROPOMINO₂ and S₂ data products in the North China Plain: Comparison, correction and application. Remote Sensing 2022, 14, 214.
Xiaoping, G.; Guangyi, L.; Yuanhang, C.; Yao, L. Research on spatiotemporal differences in pollution gas concentration in Guizhou Province based on OMI and ground monitoring. Journal of Atmospheric and Environmental Optics 2024, 19, 85-97.
Xiantong, L.; Tengfei, Z.; Qilin, W.; Haobo, T.; Xuejiao, D.; Fei, L.; Tao, D. The spatiotemporal distribution characteristics and human activity impact analysis ofNO₂ in the Pearl River Delta urban agglomeration using OMI remote sensing. Journal of Tropical Meteorology 2015, 31, 193-201

Round 2

Reviewer 1 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

1- The abstract section must be shortened to match the journal template.

2- The quality of the figures and texts in it must be revised.

3- The form of subtitles in the discussion and conclusion section is improper and does not match the journal template.

Author Response

I attach great importance to every opinion you have raised and carefully consider and respond to each one during the revision process. Your professional knowledge and experience have undoubtedly enhanced my research work. Thank you again for your hard work and valuable time on my paper. I look forward to the opportunity to continue learning and consulting from you in the future. Here is our response:

1-The abstract section must be shortened to match the journal template.

Thank you for pointing this out. We agree with this comment and have made the necessary modifications.

2-必须修改其中的图形和文本的质量。

谢谢你指出这一点。我们同意此评论并已进行必要的修改。

3- 讨论和结论部分的字幕形式不合适，与期刊模板不匹配

谢谢你指出这一点。我们同意此评论并已进行必要的修改。我们会删除不适当的副标题。

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

Comments and Suggestions for Authors

1- Why did you just focus on NO2 and didn't check other gases?

2- I cannpt be convinced that monitoring NO2 is important over mountain area. Is it possible to show the generalization of NO2 concentration over other regions like urban area?

3- Can you implement the whole process of your methodology in GEE and add the link in the paper?

4- Which preprocessing did you apply on your data like gap fiilling or masking?

5- Add quantitative results to the Abstract and Conclusion section.

6- Please check the template of this journal specially in the Reference section.