Mapping an Urban Boundary Based on Multi-Temporal Sentinel-2 and POI Data: A Case Study of Zhengzhou City

Round 1

Reviewer 1 Report

This is a generally interesting paper. My only question/comment is that while several cities are assessed in Figure 11 and Table 4, only Zhengzhou is considered in detail.

Author Response

We want to thank you for the support, constructive comments and all your work done for our papers. We have considered all comments and addressed the issues in the revision, with all changes highlighted. This report presents our responses to the comments and how we have addressed them. We now resubmit our manuscript to pursue publication in Remote Sensing. We hope to get your approval！

Author Response File: Author Response.docx

Reviewer 2 Report

This paper provides a useful contribution to the scientific debate on the delimitation of urban areas, with special emphasis on mid-scale mapping -the most commonly used by urban planners. The study takes place in China, a country where urban development issues may be critical in a number of places. The sustained pace of urban growth leads to a constant redefinition of “urban areas” because of rapidly moving external boundaries and major encroachments on neighboring natural/agricultural areas. Accordingly, the choice of Zhengzhou is a very good one.

The article is well organized and clearly written. Tables and figures are of good quality and suffer no major criticism. The author(s) present the main issues at stake, with relevant bibliographic references as a background material and the information provided ensures reproducible measurement results. Furthermore, the issue relates perfectly with the thematic area of Remote Sensing.

Specifically, I consider that the technical part dedicated to data processing poses no particular problem. The authors clearly demonstrate the efficiency of RF and machine learning in determining the extent of urban areas and boundaries. They underscore the limitations faced with the high number of training samples and verification samples. In my opinion, further discussion should be engaged on concepts rather than methods, which could lead to clarifications on a number of points.

1-As a general remark, boundaries are relative objects, e.g. they refer to objects whose identification is obtained at a given scale through dedicated sources (resolutions). The same “objects” change boundaries when they are studied at different scales (from macro to micro). By contrast, here the authors use an all-inclusive monolithic concept of boundary. They suppose that boundaries are trans-scalar object. They are supposed to overlap when determined at different scale levels -in practice, they do not. Variations occur, not only in location but also in form, especially when scale discrepancies are significant. The issue is widely discussed in the area of fractals.

2-Here, in my opinion, the authors adopt a unified conception of what boundaries are or should be. They are perfectly in line with the way urban vs. rural areas are thought of by the Chinese authorities, especially in planning terms: cities are determined by their legal/administrative status on a binary mode (corresponding to the dual-track tenure system). Binary thinking is found in the way urban and non-urban areas are adopted as exclusive categories during the RF processing. Intermediate, hybrid, fuzzy spaces are eradicated from the analysis.

3-As a consequence, surprisingly, cities are seen as absolute entities, hence the absence of any “rural” vs. “agricultural” vs. “peri-urban” areas in their perspective. Although Chinese cities expand on their outer face on rural areas with or against “village” communities (complex relationship), built-up “village” areas are considered to be “urban” without further distinction. This is confusing for foreign observers because the process of urban spread is almost lost. Cities are not only defined by external boundaries/borders, they develop through a series of complex boundaries, more or less visible, that randomly overlap because these boundaries depend on the scale/point of view adopted by the observers/actors of urban development. What about the different forms of illegal land occupation, closely connected with this study? What about rural settlements converted into “urban areas” through the expansion of state ownership?

4-As a consequence again, I consider that the proposed analysis is valid up to point 4.4: the mathematical morphology section operates a voluntary degradation of results through a simplification process validating ex-post the conceptual binary stance adopted from the very start. These operations are aimed at ensuring the replicability of results at other scale levels and this is a mistake. The adoption of a 13x13 window means that the results achieved are more consistent when a 130mx130m resolution is adopted, which makes no sense when using Sentinel-2 databases. This might be interpreted as an admission of failure. Complexity should not be sacrificed on the altar of a supposed consistency.

Other miscellaneous tips to consider:

1-lines 108-109: contradicts the general reasoning.

2-lines 109-114: not clear. Which definition/terminology is finally adopted?

3-lines 153-154: the general logic behind this multi-date choice is not explained. Isn’t it an “instantaneous” yearly analysis?

4-lines 199-201: the choice of POI variables is not really explained. Are these the most appropriate variables to introduce/describe “urban” entities?

5-fig.4: the “urban” category is still not defined as a clear concept. Cf. lines 229-231: “urban” is a result of image interpretation…

6-section 3.3.1: general principles are presented, but nothing is said about the algorithms applied.

7-line 343: disorder and fragmentation are fully operational results to be managed, not eliminated

8-lines 346-349: boundaries are not only external, they are internal too. And this is not a bias.

9-lines 355-361: in my opinion, this goes too far. The actual results are degraded on purpose to match with other (unwritten) expected results. The ambition of the authors is to provide an all-embracing/integrated/consistent map of "urban bodies".

10-fig.11: central “visual interpretation” maps are disputable. Such an introduction of subjective elements is detrimental to the demonstration. Why not overlaying boundaries for comparison?

Author Response

We want to thank you for the support, constructive comments and all your work done for our paper. You have read our article very carefully and explored many key issues in depth. We admire your exploration and rigorous spirit。 We have considered all comments and addressed the issues in the revision, with all changes highlighted. This report presents our responses to the comments and how we have addressed them. We now resubmit our manuscript to pursue publication in Remote Sensing. We hope to get your approval.

Author Response File: Author Response.docx

Reviewer 3 Report

Mapping urban boundary based on multi-temporal Sentinel-2 and POI data: A Case Study of Zhengzhou City

Addressing the following comments can help improve the quality of the manuscript

Abstract

• Line 24: define POI here at first mention
• Line 27: It is better to say: This study uses multi-temporal….

1. Introduction 

• It is vital to provide some statistics about the extent of urbanization and urban growth in China.
• What do you mean by the low quality of urbanization development and unreasonable land use planning (line 49)
• There is very little literature about the benefits of curtailing sprawl, which includes low costs for providing and maintaining infrastructure, better access to urban services, more social interaction, and walkability
• What studies used similar methods and the limitations of the studies (knowledge gap), and how does the present study intend to fill the gap?

2. Study Area and Data Sources

· What is the justification for selecting Zhengzhou as a study case?

· Figure 1 should be labeled a, b, and c to distinctly identify the country, region, and study area. 

· A link (URL) to the data source should be provided (line 149). 

4. Results 

· The section should underscore the most significant (new) findings (at the end of the section.

· Also, the findings should be linked to the research aim, stressing the value the paper adds to the literature. 

5. Discussion 

· The discussion section did not cite even one piece of literature.

· The section should situate the findings within the existing knowledge by showing where they corroborate or differ from prior studies and likely explanations.

· Mention what the results generalizable, that is, if can they be applied more broadly? 

6. Conclusion 

· There is a need to indicate the methodological implications of the findings 

· Also, highlight some implications of the study to the field of urban planning/development and management. 

Author Response

We want to thank you for the support and constructive comments. We have considered all comments and addressed the issues in the revision, with all changes highlighted. This report presents our responses to the comments and how we have addressed them. We now resubmit our manuscript to pursue publication in Remote Sensing. We hope to get your approval.

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

The following issues are still not addressed

• There is still the need to stress in the introduction the novelty of the study
• The conclusion should provide the implications of the study on urban management/development

Author Response

On behalf of all the authors of this paper, I would like to thank you for your valuable suggestions and unremitting efforts to further improve the level of our paper. Your suggestions really helped us a lot. Thank you again.

Author Response File: Author Response.docx

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

Authors adequately addresses all of my comments and queries and results in an interesting and informative paper.

Author Response

We want to thank you for the support and constructive comments. We have considered all comments and addressed the issues in the revision, with all changes highlighted. This report presents our responses to the comments and how we have addressed them. We now resubmit our manuscript to pursue publication in Remote Sensing.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear authors, I appreciate that you significantly changed and improved the article. However, just a few of the observations  I made for the previous version of the article were made. Many other observations were not taking in consideration without explanations. Since after the revision the readability of the paper was not improved and critical discussion of the results remains weak, I recommend to reject the article

Author Response

We want to thank you for the support and constructive comments. We have considered all comments and addressed the issues in the revision, with all changes highlighted. This report presents our responses to the comments and how we have addressed them. We now resubmit our manuscript to pursue publication in Remote Sensing.

Author Response File: Author Response.docx

Reviewer 3 Report

Thank you and your colleagues for the modifications that you have made to this manuscript. However, the discussion of the manuscript is still absent.

A second note: please change the colour of urban areas (in figure 5 and 10) from green to red (for instance) (red it is the unanimously accepted colour in the literature to map urban areas.

Author Response

We want to thank you for the support and constructive comments. We have considered all comments and addressed the issues in the revision, with all changes highlighted. This report presents our responses to the comments and how we have addressed them. We now resubmit our manuscript to pursue publication in Remote Sensing.

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

Thank you and your colleagues for the modifications that you have made to this article and how well you have responded to the suggestions.