Use of Landsat Imagery Time-Series and Random Forests Classifier to Reconstruct Eelgrass Bed Distribution Maps in Eeyou Istchee
Priscilla Minotti
Round 1
Reviewer 1 Report (Previous Reviewer 2)
Comments and Suggestions for AuthorsThe paper primarily discusses the extraction methods of eelgrass beds based on Landsat images, and the distribution of eelgrass beds is indeed worth studying. However, there are still many uncertainties in the extraction methods and discussions presented in the paper. Here are some suggestions:
1. First, increase the discussion on the characteristic variables of eelgrass beds. It should not be a simple listing of related vegetation indices, but rather a discussion on why these parameters can be used to improve the accuracy of eelgrass bed extraction analysis.
2. The paper mentions that total suspended solids can lead to a decrease in accuracy. The discussion section should focus on analyzing how the accuracy is affected and to what extent it might be impacted.
3. In the discussion section, the authors analyze the sources of sediment, which seems relatively unnecessary as it does not have much relevance to the main content of the paper.
4. Regarding the verification process, the comparison with aerial images is mentioned, but the paper does not describe the extraction accuracy and methods of eelgrass beds from the aerial images. It is recommended to supplement this information.
Author Response
Please see the enclosed file
Author Response File: 
Author Response.docx
Reviewer 2 Report (New Reviewer)
Comments and Suggestions for AuthorsThe study is really interesting and needs some extra reorganization so it can be followed easily.
A commented version of the manuscript is attached with suggestions to improve the organization.
Comments for author File: 
Comments.pdf
Author Response
please see enclosed file
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
Comments and Suggestions for AuthorsThe paper is generally well written but could use some improvements and clarification.
The introduction is missing the broader picture and focuses in quickly on the specific site. I think expanding the background on seagrasses more generally and their importance and how remote sensing helps with seagrass research and monitoring is needed. I think a lot of what is in the introduction should go into the section on study area with a more detailed map (including locations of dams, bays that are mentioned in the paper and other impacts) to give the readers a clearer picture of what exists there and the issues that exist.
The methods need clarification in some areas, particularly on the accuracy assessments. See detailed comments and edits in the document.
For the results, there are a lot of figures and many of them are not that useful at least not in the state they are in, i.e. the turbidity comparisons would be difficult to see for the untrained eye. I suggest applying the classification to these images so it is easier to see differences in the number of turbidity class pixels. This could then be shown in a table or graph the differences between before and after fire and maybe just an overall map showing the sites where these comparisons were made. The Landsat classifications to aerial imagery delineations are also very difficult to compare as shown. I made some suggestions in the document itself and the procedure for delineating the aerials needs to be made clear as from the imagery, borders between the eel grass vs no eelgrass are not clear at all and that seems apparent when looking at the Landsat classifications as well that often exceed those boundaries.
The discussion can use more potential explanations for the results you got. See the text as well for more details on this.
Comments for author File: Comments.pdf
The writing was overall very good. I just corrected a few grammatical issues and commented on parts that were confusing.
Reviewer 2 Report
Comments and Suggestions for AuthorsThis article mainly identifies and analyzes eelgrass beds, and the topic of the paper is of great significance, providing technical methods for identifying and changing of eelgrass beds. However, there are some issues with the article:
1. The introduction section of the paper lacks a comprehensive description of the relevant papers on the extraction of eelgrass beds. Due to the most of the eelgrass bed is under the water, the optical characteristics of water bodies have a significant impact on the recognition of submerged vegetation. Therefore, a detailed explanation of the research in this section should be provided.
2. For the classification types in the paper, whether we should consider the impact of other vegetation types, such as mangrove areas, besides eelgrass beds, and how to distinguish them from eelgrass beds.
3. All the graphic parts in the article are not very clear and do not comply with the drafting standards. It is necessary to add a north arrow, a scale, and the legend part is recommended to be included on each page to help readers better read the images. The legend in Figure 5 is not in English. The classification image in Figure 5 has no legend.
4. For the comparison with aerial images, it is recommended not only to compare the images, but also to include data result comparison, with a particular focus on describing the areas of differences. Can some verification point names be added to Figure 5 for readers to compare.
5. In the discussion section, the relationship between eelgrass beds and wildfires was discussed, but it was not discussed in depth. We cannot just show the wildfire area without describing the relationship between eelgrass beds and wildfire.
6. In the discussion section, the main features and imaging factors of eelgrass bed extraction should be discussed, and how to consider the impact of these factors, such as turbidity.
Reviewer 3 Report
Comments and Suggestions for AuthorsThis MS used Landat image to analyse the change in the area of sea grass bed in USA. The MS is straight forward and there are some minor revision suggestions:
1) Methods - please state more detail how to collect images acquired from the United States Geological Survey’s (USGS) EarthExplorer website. For non experts, it seems it need a detail procedure to indicate how to operate the website to obtain those images. May be a screen capture step by step process how to select and obtain images is needed in the supplementary materials.
2) Figure 2 legends - need further information. What is those green colour. How to highlight them and how to overlay with google earth image? Need more explanation in the legend.
3) Section 3.3 the authors stated "The area of eelgrass beds mapped with the classified images over each zone was quantified (Table 9) and plotted (Figure 7). This figure shows a de-clining trend in eelgrass extent observed throughout the study period for both zones as well as the total study period, although the 2019 extent was slightly higher than the 1991 and 1996 extents" The plot in figure 7 - are there any correlation test to test the significance of area of sea grass and year? If it is not significant in correlation analysis, no regression should be conducted. I suspect the dotted lines in figure 7 is regression line? Is this regression significant? How is the r square and p-value? This need to be clarified. It is very confused whether the relationship between the sea grass area and time is significant.
Comments on the Quality of English LanguageIt is OK for the English of the MS.
