Unravelling the Characteristics of Microhabitat Alterations in Floodplain Inundated Areas Based on High-Resolution UAV Imagery and Remote Sensing: A Case Study in Jingjiang, Yangtze River

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study analyzes the significant reduction of the floodplain area and fish microhabitats in the Jingjiang reach of the Yangtze River due to human activities and reservoir construction, utilizing high-resolution UAV images and an enhanced U-Net model for classification. The overall paper is good, but I believe the following issues need to be addressed before formal publication:

L17: 1. In the article, is it referred to as Jingjiang Reach or Jingjiang reach? I have seen studies using the first expression.
L215: I suggest providing a dedicated section in the methods to explain how the UAV data was processed in more detail, as this is an important aspect for a journal specializing in UAV technology.
L321: I think the accuracy and errors of the post-processed data should be presented in the results section, including both the UAV data and satellite remote sensing data.
L364: Please increase the font size.
L391: The same applies to this figure.
L495: The prediction of grassland coverage in the results seems to be inaccurate. What could be the reasons for this?

Author Response

Please check the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

1.    After the operation of the Three Gorges Reservoir, discharge with rainfall amount must follow its own operation manual. Hence, the flowing discharge is always less than that before operation of reservoir resulting in less water area before 2002. It would be better if annual water level and annual variation of water level could be separated into two parts of before and after operation of reservoir for regression analysis in Figure 4.
2.    After 2003, the annual variation of water level at ZC station is stable and smaller than that at CLJ station because there is regulated by reservoir while the variation of discharge from East Dongting Lake increases the amplitude of variation at CLJ station.
3.    Following the operation manual of reservoir, the occurrence frequency of inundated floodplain would be less than that before operation of reservoir. New embankment and some human-induced disturbances are naturally existed some distance from the boundary of floodplain and water zone. Except construction of port, this kind of disturbance would not significantly devastate the habitat of juvenile fish. The riparian vegetation is not included into the types of features in the JJR floodplain inundated area so that the discussion about degradation of riparian vegetation is not explored in this study.    
4.    The oscillation of validation accuracy curve in Figure 9a and validation loss curve in figure 9b are encouraged to discuss more detail.
5.    The accuracy evaluation on water, grass and shrub categories in Table 2 have good performance might result from firstly using the criteria of MNDWI > NDVI and EVI < 0.1, as well as MNDWI > EVI and EVI < 0.1. 
6.    Please discuss deeply about that extracting finer-scale features results in performance of U-Net model being better than that of the enhanced U-Net model for gravel and road classification. 

Author Response

Please check the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

Dear authors, I had the task of reviewing your manuscript "Unrevealing the characteristics of microhabitat alterations in floodplain inundated areas based on high-resolution UAV imagery and remote sensing : A case study in Jingjiang, Yangtze River".  Your research is in line with contemporary needs for the conservation of natural habitats, which are increasingly threatened by the expansion of infrastructure. 

The research is interesting, the methods are acceptable, and the results obtained are interesting and can certainly be used for various purposes. However, there are a few suggestions to make your work even better.

I think the abstract is too long, it should be shortened and all the methods used, the results obtained, as well as the novelties in the research should be clearly highlighted.

The introduction should be a definition of the starting positions and hypotheses that are the function of your research. This form and content of the introduction is too general, and the story about the importance and preservation of wetlands is redundant.

In the subheading 3.2 Modeling approach, you explain the use and purpose of the individual models you used. I suggest that you specifically emphasize in the text all the innovations you implemented in your research that relate to the models. Of course, it is necessary to present the data from testing and validating the modified models.

The Results and Discussion are well-written and I have no suggestions in this part of the text

I think you could have written the conclusion much better. You did a lot in your research, you changed and innovated a lot in terms of methodology and it is not good that this is not reflected in your conclusion. I suggest you supplement the conclusion and emphasize what the new methods you used in your research brought in the results.

Best regards.

 

Author Response

Please check the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

This paper has revised. I have no more comment.