Exploring the Relationships between Land Surface Temperature and Its Influencing Factors Using Multisource Spatial Big Data: A Case Study in Beijing, China

Round 1

Reviewer 1 Report

The innovation of this manuscript was to analyze the relationships between short-term activities with LST during the daytime and nighttime. It's worth exploring. Nevertheless, some issues should be solved or concerned before this manuscript can be published.

1.      What does “multisource spatial big data” refer to? According to my understanding, only the POI data of Sina Weibo belong to big data.

2.      LST is the research object of this paper. Therefore, many parts of the article should use SUHI instead of UHI.

3.      “human daily activities affect” should be “short-term human daily activities affect”. L65

4.      “In recent decades, heat waves enhance the intensity of UHI effect in megacities…” This is not accurate. Should be “In recent decades, heat waves usually enhance the intensity of UHI effect in megacities…” That means previous studies have found their relationships were complex and the heat waves do not necessarily enhance heat islands. L104

5.      The content is repetitive or redundant in several places.

6.      Should be “While LST is not the same as near-surface air temperature, the two are closely related.” L116

7.      Landscape configuration is also an important factor affecting the land surface temperature. The author doesn't mention it. L141

8.      Some studies analyzing the impact of human activities on LST should be mentioned, such as the total population, population density, nighttime light intensity, land use planning policy, industrial upgrading or transfer, energy structure adjustment, etc. L166

9.      “human daily activities affect” should be “short-term human daily activities affect”. L176

10.   What “human daily activities affect” in this article refer to specifically, please explain in detail.

11.   The format of “km2” was wrong in many places in the manuscript.

12.   What the different colors represent in Figure 2. It is suggested to draw false colour remote sensing map, topographic map, land use map, etc. directly.

13.   Quality control of surface temperature data is not done in this paper. These possible outliers may have a certain influence on the results.

14.   Still not clear why only autumn was chosen. I think there are significant diurnal variations in the intensity of human activity, regardless of the season. L255

15.   Why choose LST at these two time periods rather than the other two or all four time slots? L261

16.   POI data were divided into 3 categories. The author should introduce them further (as to what each category contains) in order to understand the short-term human activities on LST. L313

17.   The time of the various data should be stated clearly. Now they are unclear, including the building form, GDP, population, road, etc.

18.   Pairwise Pearson correlation analysis has only analyzed the relationships between two variables and omits the influences of other factors. This is likely to derive distorted or even wrong conclusions.

19.   The format of R2 was wrong in many places.

20.   What does “MAUP” mean? L461

21.   Figure 9 was fuzzy.

22.   There's an extra comma. L615

23.   Authors should add the “limitation and uncertain” or “limitation and future work” section.

24.   Describe results and discussions separately.

25.   The conclusions section should be simplified and clearer. Some contents do not belong to conclusions.

 

 

Author Response

The innovation of this manuscript was to analyze the relationships between short-term activities with LST during the daytime and nighttime. It's worth exploring. Nevertheless, some issues should be solved or concerned before this manuscript can be published.

Response: Thanks very much for your meticulous review and encouraging comments. We strive to address your concerns and hope to gain your further support.

 

1. What does “multisource spatial big data” refer to? According to my understanding, only the POI data of Sina Weibo belong to big data.

Response: Thank you for this comment. We appreciate and understand your question. In the current article, we refer Sina Weibo check-in data (acquired from mobile phone metadata), POI data, MODIS land surface temperature data, building contour vector data, 1 km Grid population and GDP data, and road network data as spatial big data. While the remote sensing and vector GIS data are usually not specifically discussed as spatial big data, at the spatial (1 km) and temporal (day and night) resolutions in our current study, we believe their volumes are sufficiently big enough to be categorized as spatial big data. We have clarified this point in the revision and hope this explanation will gain you understanding and approval.

 

2. LST is the research object of this paper. Therefore, many parts of the article should use SUHI instead of UHI.

Response: Thank you for your suggestion. We have replaced “UHI” to “SUHI” in some places to make our description more accurate and rigorous.

 

3. “human daily activities affect” should be “short-term human daily activities affect”. L65

Response:  We have followed your suggestion and modified this term here. Thank you.

 

4. “In recent decades, heat waves enhance the intensity of UHI effect in megacities…” This is not accurate. Should be “In recent decades, heat waves usually enhance the intensity of UHI effect in megacities…” That means previous studies have found their relationships were complex and the heat waves do not necessarily enhance heat islands. L104

Response: Thank you for your careful review here. We have modified this sentence according to your suggestion.

 

5. The content is repetitive or redundant in several places.

Response:  Thank you for your suggestion. We have deleted some repetitive and redundant content in Introduction section and we hope this revised version will gain your support.

 

6. Should be “While LST is not the same as near-surface air temperature, the two are closely related.” L116

Response: We have followed your suggestion and modified this sentence here. Thank you very much.

 

7. Landscape configuration is also an important factor affecting the land surface temperature. The author doesn't mention it. L141

Response: Thank you for this comment. We agree with your opinion that landscape configuration plays an important role in affecting LST. Because the study area is relatively small and buildings take up major space in the study area, we used the building forms (such as average area and volume of buildings) as representatives of landscape configuration. Though it is difficult to depict details of landscape, but the area and volume of buildings can reflect some characteristics of their configuration to a certain extent. We added this comment in the revised version. We hope our explanation will gain your support.

 

8. Some studies analyzing the impact of human activities on LST should be mentioned, such as the total population, population density, nighttime light intensity, land use planning policy, industrial upgrading or transfer, energy structure adjustment, etc. L166

Response: Thank you for your suggestion. We have cited several studies and added corresponding narratives in the article here to support our argument.

 

9. “human daily activities affect” should be “short-term human daily activities affect”. L176

Response: Modified. Thank you.

 

10. What “human daily activities affect” in this article refer to specifically, please explain in detail.

Response: Thank you for this comment. The human daily activity in this article is a relatively broad concept, we refer human daily activities as what urban residents experience and participate in a 24-hour period. They include home-based, work-based activities, commuting, shopping, dining out, and other leisure activities that happen in one day. In the current study, we argue that human activities are accompanied by an increase in heat release caused by energy consumption, so they will have an impact on LST. We added these comments to the revision. We hope this explanation will gain your understanding and support.

 

11. The format of “km2” was wrong in many places in the manuscript.

Response: Thank you for your meticulous review. We have modified in the article.

 

12. What the different colors represent in Figure 2. It is suggested to draw false colour remote sensing map, topographic map, land use map, etc. directly.

Response: Thank you for this comment. You are right that using different colors without legend introduces confusion. Since the purpose of Figure 2 is to show the location of our study area, we replaced the original color map with a simpler administrative boundary map to avoid confusion. We hope this modification will gain your approval.

 

13. Quality control of surface temperature data is not done in this paper. These possible outliers may have a certain influence on the results.

Response: Thank you for this comment. The surface temperature data is derived directly from MODIS data. From the results of descriptive statistics, we can see that the largest range of LST is 1.33 ℃, which occurs during the daytime at the pixel level. This is within a reasonable range and shows high quality of these MODIS data. We incorporate your comment to the narrative and hope our revision will gain your understanding and support.

 

14. Still not clear why only autumn was chosen. I think there are significant diurnal variations in the intensity of human activity, regardless of the season. L255

Response: Thank you for your comment. We agree that there are significant diurnal variations in intensity of human activities. According to studies cited in the manuscript, urban landscape is usually the most significant factor of LST. Since we aim to investigate the impact of short-term human daily activities on urban LST, we hope to limit other factors’ influence if possible. In both summer and winter times, the urban landscape and vegetation coverage might make human activities’ influence on LST less detectable.  We added this justification to this revision and hope to gain your understanding and support.

 

15. Why choose LST at these two time periods rather than the other two or all four time slots? L261

Response: Thank you for the comment. We added more clarification to the revision and hopefully it will address your concern: “We chose these two time slots to extract daytime and nighttime LST is because these two periods are when the LST changes might be most influenced by the change of human activities. The daytime period is right after the city “wakes up.” People commuted to work; offices were in full operation. The nighttime period is when the city “goes to sleep.” Most people went to bed, and nighttime businesses closed. During these two periods, human activities’ influence on LST are most silent.

 

16. POI data were divided into 3 categories. The author should introduce them further (as to what each category contains) in order to understand the short-term human activities on LST. L313

Response: Thank you for your suggestion. We have added a relevant narrative to improve readers’ understanding of POI data used in the current study.

 

17. The time of the various data should be stated clearly. Now they are unclear, including the building form, GDP, population, road, etc.

Response: Thank you for this comment. We have added relevant time information of various data and we also argue that although the data is not collected in the same year, the urban landscape’s influence on urban LST remains relatively stable since the central areas of Beijing was already highly developed and changes of urban landscape over the years (2014 – 2020) might cause negligible changes on its influence on urban LST. We hope this addition will gain your support.

 

18. Pairwise Pearson correlation analysis has only analyzed the relationships between two variables and omits the influences of other factors. This is likely to derive distorted or even wrong conclusions.

Response: Thank you for your insightful comment. We agree with your comment, Pearson correlation analysis can only analyze the relationship between two variables without controlling other factors. We are using the correlation analysis to obtain some intuitive and useful information about the relationship between two variables, especially in the absence of other data. We use this information as a preliminary examination of the potential relationships between the variables that are prepared for multiple-linear regression. The regression model is our tool to examine the impact of each factor on LST under the condition of controlling the influences of other factors. In order to improve the quality of regression results, stepwise regressions and spatial autoregressive model are adopted to eliminate the problem of multicollinearity and residual spatial autocorrelation. We added brief elaborations in the revision. We hope this explanation will gain your understanding and approval.

 

19. The format of R2 was wrong in many places.

Response:  Modified and thanks for your careful review.

 

20. What does “MAUP” mean? L461

Response: “MAUP” is the abbreviation of “modifiable areal unit problem”. This problem leads to some differences of the impacts of influencing factors at different spatial scales. We have added this term in the article and thank you for this comment.

 

21. Figure 9 was fuzzy.

Response: Than you for your suggestion. We have replaced Figure 9 and Figure 10 with tables to make the results of Pearson correlation analysis more direct and readable.

 

22. There's an extra comma. L615

Response: Deleted. Thanks for your careful review.

 

23. Authors should add the “limitation and uncertain” or “limitation and future work” section.

Response: Thank you for this constructive suggestion. We have adjusted the structure of the article and added “limitation and future work” section. We hope the modification would gain your approval.

 

24. Describe results and discussions separately.

Response: We have followed your suggestion and added discussion section in the article. Thank you.

 

25. The conclusions section should be simplified and clearer. Some contents do not belong to conclusions.

Response: Thank you for your suggestion. We have simplified the conclusion section and deleted some irrelevant contents to make this section more concentrated.

Reviewer 2 Report

The manuscript entitled “Does human daily activities matter? Exploring the relationships between land surface temperature and its influencing factors using multisource spatial big data” is interesting and certainly makes an important contribution to the field. However, some parts should be improved/clarified. Please find specific comments below:

 

- Please include the case study in the title.

 

- Figure 2 lacks a map of China in the Asian context.

 

- Please improve the readability of figures 4 and 6.

 

- The caption of figures 4/ 5/ 6/ 7 should contextualize the map spatially and temporally. Additionally, these maps must also have a graphic scale and geographic orientations and/or coordinates.

 

- The methodology needs more explanations regarding alternative approaches.

 

- The validation process is missing. Therefore, there is no way to effectively identify the veracity of the results.

 

- In the discussion section the authors could highlight open questions and future research directions/challenges.

 

- What are the main contributions of this study? Are they new compared to previous studies? These things must be carefully discussed (in the discussion this is done in a very superficial way). Authors should clearly demonstrate the contribution of this study to the field of knowledge. This study has to generate some kind of contribution to the scientific field (e.g., the proposal of a new framework, an innovative approach to a methodology, etc.). This point is not very clear and the authors should improve this part of the study.

 

- The authors should cover the weaknesses/ limitations of their study.

 

- Minor grammar and punctuation errors can be found throughout the text and need to be corrected.

Author Response

The manuscript entitled “Does human daily activities matter? Exploring the relationships between land surface temperature and its influencing factors using multisource spatial big data” is interesting and certainly makes an important contribution to the field. However, some parts should be improved/clarified. Please find specific comments below:

Response: Thank you very much for your meticulous review and encouraging comments. We strive to address your concerns and hope this version will gain your approval.

 

- Please include the case study in the title.

Response: Thank you for this constructive suggestion. We have changed the title to “Exploring the relationships between land surface temperature and its influencing factors using multisource spatial big data: A case study in Beijing, China”.

 

- Figure 2 lacks a map of China in the Asian context.

Response: Thank you for your suggestion. Considering that Asia map will take up too much space to focus on the study area, we just added a map of China to show the location of Beijing. We hope this revision will gain your understanding.

 

- Please improve the readability of figures 4 and 6.

Response: Thank you for this comment. Due to the coordinate system setting, the grid of LST raster data is in the shape of parallelogram, as shown in Figure 4 and 6. We have tried to transfer parallelogram to rectangle, but it is not good for the comparison with administrative maps (Figure 5 and 7). Therefore, we kept these two figures in parallelogram but added scale bar and compass to improve their readability. We hope this revision will gain your support.

 

- The caption of figures 4/ 5/ 6/ 7 should contextualize the map spatially and temporally. Additionally, these maps must also have a graphic scale and geographic orientations and/or coordinates.

Response: We have followed your suggestion and modified relevant content of Figure 4/5/67. We added study area and time to the captions and drew a scale bar and compass in each figure. Thank you for your suggestion.

 

- The methodology needs more explanations regarding alternative approaches.

Response: Thank you for your suggestion. We have added a relative narrative about the comparison between the commonly used ordinary least square model and spatial autoregressive model in the article.

 

- The validation process is missing. Therefore, there is no way to effectively identify the veracity of the results.

Response: Thank you for your insightful suggestion. We carefully discussed this issue among the authors. As the discussion result, we agree that since we are not really trying to validate our regression model, instead, the regression model is established to test the relationships in our particular case (central Beijing, China), we do not add a validation process. But we did add the Lagrange Multiplier tests to validate the veracity of the spatial model, and provide model tests to validate the results (AICs and confidence intervals) are legitmate. We hope this explanation will gain your understanding and support.

 

- In the discussion section the authors could highlight open questions and future research directions/challenges.

Response: Thank you for this constructive suggestion. We have adjusted the structure of the article and added “limitation and future work” section. We hope this modification would gain your approval.

 

- What are the main contributions of this study? Are they new compared to previous studies? These things must be carefully discussed (in the discussion this is done in a very superficial way). Authors should clearly demonstrate the contribution of this study to the field of knowledge. This study has to generate some kind of contribution to the scientific field (e.g., the proposal of a new framework, an innovative approach to a methodology, etc.). This point is not very clear and the authors should improve this part of the study.

Response: Thank you for this critical comment. As we argued in the article, the main contribution of this study is that in terms of theoretical contribution, we incorporate short-term human daily activities into the mechanism analysis of LST. Compared with previous studies, we study LST from a micro perspective that adds to the knowledge of understanding everyday urbanism and provide strategies for micromanagement of urban planning and urban sustainability. We hope our explanation will gain your support. We added these comments in the newly created discussion section. We hope the addition and modification will gain your approval and support.

 

- The authors should cover the weaknesses/ limitations of their study.

Response: We have added relevant contents in the “limitation and future work” section. Thank you.

 

- Minor grammar and punctuation errors can be found throughout the text and need to be corrected.

Response: Thank you for your careful review. We have asked a native speaker to carefully review through the manuscript to correct any possible grammatic error and wording issue. We hope this revision would gain your support.

 

Reviewer 3 Report

The authors presented an analysis of relationships between LST and social media check-ins for daytime and nighttime. The analysis revealed a potential feature of social media for explaining UHI through human activities. Since this is a practice of combining satellite remote sensing and social media, a sort of big data, the manuscript is suitable for the special issue. However, I recommend major revision before the publication process, specifically for visualizing the social media data for readers who are not familiar with such datasets. I would appreciate authors to address my comments below.

1. Please add brief analyses of the social media data, such as showing the spatial distribution of the check-ins and density. This should support readers to understand that the social media could represent human activities for explaining UHI. Preferably, spatial distribution of check-ins should be compared with the LST to explain features of social media useful to model human-induced UHI.

2. I suspect pseudo-correlation between check-ins and LST through population. Check-ins of night-time are likely arrivals at home, so it could have strong correlations between check-ins and resident population. Please consider analyzing correlations between night-time check-ins and resident population for inspecting multicollinearity in the regression analysis.

3. Figure 9 and 10 is not visible. Please replace them with tables.

4. Line 278 “Sina Weibo Check-in data records the exact time, location (in geographic coordinates), and the content of the Weibo (mini blog) when the user logged in. “: Precision of the location information should be referred from technical documents of Webo Check-in data or past studies.

5. Line 672 “First, check-in data serves as a good proxy of human daily activities. “: The study has not yet reached this conclusion because of lacking validation with ground truth. Otherwise, the author should cite results of the past studies.

 

Thank you in advance for your consideration.

Author Response

The authors presented an analysis of relationships between LST and social media check-ins for daytime and nighttime. The analysis revealed a potential feature of social media for explaining UHI through human activities. Since this is a practice of combining satellite remote sensing and social media, a sort of big data, the manuscript is suitable for the special issue. However, I recommend major revision before the publication process, specifically for visualizing the social media data for readers who are not familiar with such datasets. I would appreciate authors to address my comments below.

Response: Thank you very much for your meticulous review and encouraging comments. We strive to address your concerns and hope this revised manuscript will gain your support.

 

1. Please add brief analyses of the social media data, such as showing the spatial distribution of the check-ins and density. This should support readers to understand that the social media could represent human activities for explaining UHI. Preferably, spatial distribution of check-ins should be compared with the LST to explain features of social media useful to model human-induced UHI.

Response: Thank you for your insightful suggestion. We agree very much with your comment here. It is very tempting to map the check-ins’ spatial distribution to establish a visual link between human activities and urban land surface temperatures. As a matter of fact, one of our preliminary explorations is to map the check-in data (not separated in different time slots) and hope we can get direct visual linkage between check-ins and urban land surface temperatures. The visualization, however, does not provide clear linkage. We discussed among ourselves and agree that this is likely because the temperature changes across the study area is not varying enough to show a direct visual link between lumped check-ins and temperature variation. In addition, as argued in choosing only two time slots of the MODIS data to represent urban daytime and nighttime land surface temperatures, the influence of human activities on urban LST is quite subtle. While the statistical correlation between the two does exist, a visual link is not immediately presentable. We added this brief explanation in our revision. We hope this explanation will gain your understanding and support.

 

2. I suspect pseudo-correlation between check-ins and LST through population. Check-ins of night-time are likely arrivals at home, so it could have strong correlations between check-ins and resident population. Please consider analyzing correlations between night-time check-ins and resident population for inspecting multicollinearity in the regression analysis.

Response: Thank you for this critical comment. In the current study, we have investigated multicollinearity among the explanatory variables in the regression analysis. On one hand, we used the logarithmic transformation of variables to weaken the potential strong correlation between check-ins and resident population. On the other hand, we employed stepwise regression method to filter explanatory variables until their variance inflation factors (VIFs) are less than 10. In fact, all the VIFs of check-ins and resident population are less than 5. We have added this clarification in the revision. We hope our explanations will gain your support.

 

3. Figure 9 and 10 is not visible. Please replace them with tables.

Response: We have followed your suggestion and replaced Figure 9 and 10 with tables to make the results of Pearson correlation analysis more direct and readable. Thank you.

 

4. Line 278 “Sina Weibo Check-in data records the exact time, location (in geographic coordinates), and the content of the Weibo (mini blog) when the user logged in. “: Precision of the location information should be referred from technical documents of Webo Check-in data or past studies.

Response: Thank you for this comment. We have carefully reviewed some past studies that use Weibo check-in data (such as 1. Rizwan, M., et al. (2018). "Using Location-Based Social Media Data to Observe Check-In Behavior and Gender Difference: Bringing Weibo Data into Play." Isprs International Journal of Geo-Information 7(5); 2. Wu, C., et al. (2018). "Check-in behaviour and spatio-temporal vibrancy: An exploratory analysis in Shenzhen, China." Cities 77: 104-116; and 3. Tu, W., et al. (2020). "Portraying the spatial dynamics of urban vibrancy using multisource urban big data." Computers Environment and Urban Systems 80.). The location information of Weibo check-in data was found to be precise and reliable. We have added the relevant citations. In addition, we examined the precision of their location information by overlaying administrative polygon map, road network line map and check-in data point map. We found that the locations of check-in points are mainly distributed on both sides of the road. This distribution pattern verified the precision of the location information to a certain extent as well. We have added this brief clarification in our revision. We hope our explanations will gain your understanding and approval.

 

5. Line 672 “First, check-in data serves as a good proxy of human daily activities. “: The study has not yet reached this conclusion because of lacking validation with ground truth. Otherwise, the author should cite results of the past studies.

Response: Thanks very much for your suggestion. We have revised inaccurate narratives in this section and added relevant citations to support our argument here as well. We hope this version will gain your support.

 

Thank you in advance for your consideration.

Response: Thank you for your meticulous review.

Reviewer 4 Report

Thank you for this paper - I think overall the concept makes sense, however I have a lot of concerns about the data used in Section 3.3:

-Is the data obtained from Sina Weibo available for everyone worldwide to use?

-Is Baidu POI data accessible to researchers outside of China?

Also - overall I didn't see any mention of what software/hardware configuration was used to conduct the research.

 

I think these questions would need to be answered before accepting for publishing.

 

 

Author Response

Thank you for this paper - I think overall the concept makes sense, however I have a lot of concerns about the data used in Section 3.3:

Response: Thank you for your encouragement and comments. We strive to address all your concerns fully.

-Is the data obtained from Sina Weibo available for everyone worldwide to use?

Response: While the data was collected from within China, a quick google search pointed to this website: http://www.cs.cmu.edu/~lingwang/weiboguide/ that provides very detailed guidance of how to use the Sina Weibo API to acquire Sina Weibo data. Still, access to the data may be restricted or limited for non-Chinese users or entities. We added a brief statement of this restriction in our revision, and hope this will address your concern.

-Is Baidu POI data accessible to researchers outside of China?

Response: Thank you for this informative question. Again, a quick google search pointed to this website: https://github.com/wangunji/POI-data-of-China/, which contains only four cities’ POI data in 2020 obtained from Baidu Map. From our further exploration, we found that Baidu does offer an application programming interface (API) that allows developers and researchers to access some of the data, subject to certain terms and conditions. These terms and conditions may include restrictions on the use of the data, as well as requirements to obtain permission from Baidu before using the data for certain purposes. Based on this information, we believe the Baidu POI data is partially accessible to researchers outside of China. We added a brief statement of this restriction in our revision, and hope this will address your concern.

Also - overall I didn't see any mention of what software/hardware configuration was used to conduct the research.

 Response: Thank you for this careful review. We have added the software packages used in the calibration of our model and carrying out the tests and analyses: “All the analyses and tests are conducted in the statistical software platform R with the packages spdep and spatialreg.” Relevant citations referring to the software platform and packages are also included. We hope this addition would gain your approval and support.

I think these questions would need to be answered before accepting for publishing.

 Response: Thank you very much for your support. We hope our revision addresses your concerns fully.

Round 2

Reviewer 1 Report

       I'm glad to see that the paper has been improved to a greater extent. But there are still a few requests, concerns or suggestions.

1. In my opinion, the urban area of Beijing city center is large enough to analyze the influence factors of LSTs. Moreover,

It is not difficult to obtain the landscape configuration of Beijing on the two-dimensional level. In the last decade, several people have done related research. It is indeed difficult to quantify three-dimensional landscape configuration, and few reports have been seen. Related descriptions can be placed in the limitations and future work section.

2. The location map alone cannot give the reader an adequate understanding of the general situation of the study area. Therefore, I still want the authors to add the false colour remote sensing map, topographic map, land use map, etc. directly. It's not hard to do this work and it doesn't take much time.

       3. The authors thinks that the urban landscape and vegetation coverage might make human activities’ influence on LST less detectable in both summer and winter times. There are no references here. This idea may be the author's subjective inference. The author only analyzes autumn and does not analyze summer and winter. This may be a shortcoming and future work. The authors can add related descriptions in the limitation and future work section.

    4. “Figure 9 is fuzzy” means figure 9 is not clear enough and the resolution is not high. It is recommended to keep Figure 9 and not use a table.

    5. The conclusion section was not significantly revised. There is no significant change compared to before.

Author Response

  I'm glad to see that the paper has been improved to a greater extent. But there are still a few requests, concerns or suggestions.

Response: Thank you very much for your meticulous review again. We strive to address your concerns and hope to gain your further approval.

 

1. In my opinion, the urban area of Beijing city center is large enough to analyze the influence factors of LSTs. Moreover, it is not difficult to obtain the landscape configuration of Beijing on the two-dimensional level. In the last decade, several people have done related research. It is indeed difficult to quantify three-dimensional landscape configuration, and few reports have been seen. Related descriptions can be placed in the limitations and future work section.

Response: Thanks for your insightful suggestion. We have discussed this issue in the limitations and future work section. We hope this modification will gain your support.

 

2. The location map alone cannot give the reader an adequate understanding of the general situation of the study area. Therefore, I still want the authors to add the false colour remote sensing map, topographic map, land use map, etc. directly. It's not hard to do this work and it doesn't take much time.

Response: Thank you for helping us get the manuscript better. We replaced the location map with a land use map, with the different land use types colorized. We hope this modification will gain your support.

 

3. The authors thinks that the urban landscape and vegetation coverage might make human activities’ influence on LST less detectable in both summer and winter times. There are no references here. This idea may be the author's subjective inference. The author only analyzes autumn and does not analyze summer and winter. This may be a shortcoming and future work. The authors can add related descriptions in the limitation and future work section.

Response: Thanks for your suggestion. We have added related descriptions in the limitation and future work section. We hope this addition will gain your support.

 

4. “Figure 9 is fuzzy” means figure 9 is not clear enough and the resolution is not high. It is recommended to keep Figure 9 and not use a table.

Response: Figure 9 and Figure 10 have been replaced by more figures (Figures 9 to 18) to show the results more clearly. Thank you very much.

 

5. The conclusion section was not significantly revised. There is no significant change compared to before.

Response: Thanks for your suggestion. We have revised the conclusion section carefully again. Some irrelevant contents are deleted, and some narratives are added to make our conclusions more concise. We hope this modification will gain your support.

 

We sincerely thank you for your many helpful comments and meticulous review, which help us to make the manuscript better.

Reviewer 2 Report

The authors answered all my concerns. Thus, from my point of view, the manuscript may be accepted for publication.

Author Response

The authors answered all my concerns. Thus, from my point of view, the manuscript may be accepted for publication.

Response: Thank you for your comments.

Reviewer 3 Report

Thank you for the responses. I confirmed my comments are addresssed.

Author Response

Thank you for the responses. I confirmed my comments are addresssed.

Response: Thank you for your support.

Reviewer 4 Report

Thank you for providing responses and adding in detail in your manuscript:

- Science is intended to be reproducible - if data access is restricted for outside of China users, what alternative datasets other than Sina Weibo and Baidu POI can be used to closely reproduce your research for Beijing specifically? If it is not possible to get this easily outside of China - or something similar - then please make a note of this. 

- While the mention of the software packages was useful - could the hardware configuration be also mentioned so that the reader can identify what type of machine and resource allocations they would need to reproduce your research?

Thank you.

Author Response

Thank you for providing responses and adding in detail in your manuscript:

Response: Thank you very much for your approval.

 

- Science is intended to be reproducible - if data access is restricted for outside of China users, what alternative datasets other than Sina Weibo and Baidu POI can be used to closely reproduce your research for Beijing specifically? If it is not possible to get this easily outside of China - or something similar - then please make a note of this.

Response: Thank you for this comment. We have made a note and commented that similar data type such as geotagged twitter and Facebook data can be used for this type of research.  POI data can also be obtained from other platforms, such as OpenStreetMap. We made relevant notes in the revision and hope this explanation will gain your support.

 

- While the mention of the software packages was useful - could the hardware configuration be also mentioned so that the reader can identify what type of machine and resource allocations they would need to reproduce your research?

Response: Thank you for this comment. In the current study, our computer is equipped with Intel Core i5 8^th CPU and 8 GB memory. We added this information in the revision, and we hope the information will help the readers for better management of their own analysis with similar data types.

Thank you very much for your helpful comments and review.