A Structure Identification Method for Urban Agglomeration Based on Nighttime Light Data and Railway Data

Round 1

Reviewer 1 Report

This paper uses the nighttime light data and railway operation data to obtain the urban agglomerations with the community detection Louvain algorithm. Considering the contribution and methodology, I do not think it is qualified to be published in Remote Sensing. Here are the detailed suggestions.

 

1. The contribution of this work is not clear. There are many works using multisource data to delineate urban agglomerations (He et al., 2021, Remote Sensing; Cao et al., 2020, Remote Sensing of Environment). If the contribution of this work is about the method, the Louvain algorithm is not new; if the contribution lies in the application, then the data used (year 2013 and 2014) might seem a little outdated. These two kinds of datasets could be easily updated.

 

2. The methodology of this work I suppose is so sound. Cities are developing, there is no truth value for urban agglomeration. It might not be reasonable to calculate the “accuracy of urban agglomerations identification” by using the detected sum of areas divided by the reference data. We could compare the observed results with city development guidelines to see whether cities are agglomerated crossing the pre-defined boundaries.

 

3. Another questionable point is about deriving nighttime light urban network through the Guass attenuation function. The nighttime light data could delineate urban built-up areas, but deriving city connections only through the distance between them (Equation 1 in the paper) is not scientific. It is the same as using the distance network to derive city connection strength in essence, then the use of nighttime light data seems meaningless. There are many kinds of city connections, why not choose other kinds of relations.

 

4. I do not understand why the study area is confined to the eastern part of Hu Line. Even if the guides about urban agglomerations are all in the eastern part, the authors should also examine the current situation in the whole China. There is not opposed to the studies focusing on one specific urban agglomeration.

 

5. The authors note that “the adjusted railway data has 269 noes and 648 lines” (Line 200). The data volume might seem a little small for the area.

 

6. There are many typos in the paper. “are common data source”(Line 61), “ is regarded” (Line 64), “consider” (Line 101)

 

7. The citing and the reference styles are not right.

Comments for author File: Comments.pdf

Author Response

请参阅附件。

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear Authors

The manuscript presents a method for the identification of spatial structures for urban agglomerations supported by heterogeneous data and a complex network.

The adopted approach of dividing the built-up urban areas allowed their analysis by resorting to Defense Meteorological Satellite data, and in urban objects to obtain the Urban Network by municipal administrative region.

The text is well structured and the methodology and discussion present a solid basis, so I believe that the text can be published in its present form.

However, I emphasize that there is an evident limitation in terms of the bibliography with citations that are not comprehensive in terms of thematics, however it is not advisable to impose specific authors to be cited.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The contribution is interesting and original because it proposes an approach, the Composite Urban Network, that could be scalable and replicable in other cities. The literature is coherent particularly the one on methods is very comprehensive. I would suggest expanding the discussion on the entire research path by highlighting the advantages and limitations of the whole methodological approach.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Accept.