Contribution of Land Cover Classification Results Based on Sentinel-1 and 2 to the Accreditation of Wetland Cities
, Yawen Deng 1,2, Xiaogan Yin 1,2, Kaifeng Peng 3, Pinzeng Rao 4,5 and Zhuo Li 1,2Round 1
Reviewer 1 Report
This is a potentially interesting submission that uses Sentinel-1 and 2 data to classify the land cover of the first six wetland cities recognized in China. The topic has merit and seems to fit well with the scope of the journal. The article is easy to read and gives a well-presented overview of change in the study area. There are some minor concerns as follows:
1. Please check the title again, there is a space missing between “1” and “and”.
2. Because the classification method is an important basis for extracting the spatial distribution of wetlands, it is suggested to make a proper description and summary in the introduction, especially in lines 108-110. Why choose the random forest method?
3. It is very important to establish a scientific and reasonable land cover classification system. It is suggested to add a reference to Table 2 or explain the reasons.
4. Figure 2 is not very clear. It is recommended to delete the shadow in the figure or change it to a line chart. There are similar problems in figures 9 and 10.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This paper takes the first batch of wetland cities in China as the research areas for the first time. It obtained 10 m land cover products and calculated the wetland rate and wetland protection rate to support the creation of wetland cities. The surface types and wetland conditions of wetland cities are analyzed in detail. Meanwhile, the experimental results and the analysis make the work persuasive. Some minor grammatical improvements could be made to improve the readability of the paper. Despite this, I think this paper is a good contribution and recommend this manuscript for publication and pending minor revisions as listed below.
1. Line 77. “km2” > “km2” 2. Line 111 to 121. It is suggested that the content of this study should be expressed more clearly. What is the purpose of the research? What are the contents of the research? What is the significance of the research? 3. Line 125/286/298/305/328. “Harbin”> “Haerbin” 4. 2.1 Study Area. What is the source of water depth data, it is recommended to add in the data section.Haikou and Dongying are coastal cities and have coastal extension areas. Please clarify the area of administrative region and coastal extension regions in Haikou and Dongying in Table1. 5. 2.3.1 Classification Scheme. Does the land cover classification scheme in Table 2 have a reference basis? It is recommended to add 6. Line 227. 2.3.6 Calculate the WR and WPR to support wetland city accreditation. The abbreviation of WR and WPR are used for the first time, add the full name. 7. Line 284 to 285. “abundant” > “common” 8. The legend of Figure 9 and Figure 10 “Figure 9. Wetland rate for 6 cities in 2015 and 2020.” > “Figure 9. Wetland rate (WR) for 6 cities in 2015 and 2020.” “Figure 10. Wetland protection rate for 6 cities in 2015 and 2020.” > “Figure 10. Wetland protection rate (WPR) for 6 cities in 2015 and 2020.” 9. Line 379. “Figure 11. Wetland proportion for wetland types for 6 cities in 2015 and 2020.” > “Figure 11. Wetland proportion for wetland types for 6 cities in 2015 (a) and 2020 (b).” 10. Line 381 to 384. This is the introduction of the template, please delete. 11. Line 385. 4.1 Advantages of This Study. This expression is a bit strange, it is recommended to change it to “Comparison with Other Studies” and increase the comparison with other urban wetland research. 12. 4.2 Reasons for the Change in Wetland Rate and Wetland Protection Rate in Different Cities. It is mentioned in 4.2 that the decline in the wetland protection rate of Haikou and Dongying may be related to the coastline. Can you add the image comparison between 2015 and 2020 to intuitively illustrate this situation?Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The paper aimed to demonstrate the effectiveness of the Sentinel image archive and RF application for a detailed wetland assessment. Some issues need to be addressed for publication.
1. L137: Why did the authors choose 2015 despite the disadvantage of fewer Sentinel-2 images for the year?
2. L142: Sentinel-2 temporal resolution is 10 days? Sentinel-1 information is missing.
3. 2.3.1 I think the classification here is crucial if the study aims to contribute to the worldwide holist assessment. It needs to show a more solid grounded explanation for determining the land classes to be identified.
4. 2.3.2 All the samples were used simultaneously? If the study aims to contribute to worldwide holist assessment, it might be worth comparing RF classifiers created from each city data set.
5. L197-200: The sampling approach needs to be briefly explained as it is essential information.
6. 2.3.3 Used bands and indexes need to be explained as it is essential information.
7. L210-211: The 10-fold random forest algorithm needs to be briefly explained as it is essential information. Also, what types were identified in each iteration should be provided (maybe in Appendix) to show how the results were robust.
8. Figure 5: More detailed assessment would enrich the paper's contribution, such as which land cover class is often misclassified in which land cover map. The differences in categories are not necessarily essential information when discussing which land cover map is "better." This point is related to the above comment 3.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 3 Report
I found the comments were mostly addressed.
But, regarding the satellite image descriptions, Sentinel-2's temporal resolution is five days. A reference for this kind of information should be from the authorities' websites or reports but not from an ordinary single article. In addition, the number of used images from each satellite archive and the acquisition date should be mentioned. It is standard practice.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
