A Robust Index Based on Phenological Features to Extract Sugarcane from Multisource Remote Sensing Data

Round 1

Reviewer 1 Report

you need to add and clarify the following remarks in your work: 

Did the study previously depend on gathering comprehensive and representative training samples, making the mapping of large sugarcane fields time-consuming and inefficient?

Is the study's methodology solely based on the use of the Normalized Difference Vegetation Index (NDVI) Based Sugarcane Index (NBSI) for sugarcane extraction, potentially limiting the scope of the research?

Are the threshold settings for NBSI considered as simple, and does it rely on these thresholds for mapping sugarcane cultivation areas?

Is there no detailed discussion on the disadvantages or limitations of the NBSI method?

Potential Improvements:

Could the study benefit from exploring more advanced and diversified methodologies beyond NBSI to provide a more comprehensive analysis of sugarcane mapping?

Enhance the current state of knowledge about Support Vector Machines (SVM) by including illustrative examples: 

https://doi.org/10.1117/12.2028640

10.1109/MVA.2015.7153187

https://doi.org/10.3390/rs13204040

https://doi.org/10.1117/1.JEI.28.2.021012

Is it possible to refine the threshold settings used in NBSI to further enhance its accuracy and applicability in different regions?

Can the study consider a more extensive and nuanced analysis of the advantages and disadvantages of the NBSI method?

Are there opportunities to incorporate more sources of data or additional remote sensing techniques to further improve the accuracy and robustness of sugarcane mapping?

Could the research expand beyond a single region, potentially including a broader range of locations and conditions for a more comprehensive evaluation of NBSI's performance?

Need to be improved.

Author Response

请参阅附件。

Author Response File: Author Response.docx

Reviewer 2 Report

Sugarcane is vital for both sugar and biofuel production, making it an essential contributor to the strength of the sugar industry. This manuscript reported a new novel Sugarcane Index called NBSI derived from combining imagery from MODIS, Sentinel-2, and Landsat-8, which was utilized for sugarcane mapping. This study found that the NBSI was highly effective in mapping sugarcane cultivation areas using simple threshold settings and achieved a superior accuracy with an overall accuracy (OA) of 95.24% and a Kappa coefficient of 0.93 compared to SVM and RF. The findings give an insight into accurately and effectively mapping sugarcane. The study might be interesting to Remote Sensing readership, but still required potential improvements before being published. My major concerns include:

Major comments

1)      The NBSI was developed utilizing NDVI data from multiple remote imagery sources in 2021, but its robustness has not been evaluated using the latest imagery. In addition, the NBSI includes four measurement indicators: f(W1 ),f(W2 ),f(V) and f(D), which means that the NBSI might not accurately delineate sugarcane cultivation in the early growth stage of sugarcane due to the missing of f(V) and f(D).

2)      Although the NBSI outperformed RF and SVM for sugarcane mapping, this manuscript did not provide detailed information regarding the specific vegetation indices, textural features, and canopy cover used to extract sugarcane for RF and SVM. It is unclear whether the comparison with NBSI involved sugarcane mapping of RF and SVM for the entire sugarcane growth period or individual growth periods.

Detailed comments

1）Page 2 Lines 66, the citation format of Jiang Hao[25] et al. should be Jiang Hao et al. [25]. Please check the entire manuscript.

2）Page 9, As for Figure 5, captions should be self-explanatory. For each figure and table, there should be clear descriptions and interpretations, allowing the reader to understand the significance of the data without referring to the text. Therefore, all captions in the paper should be enlarged to explain much better what is contained in the figure or table without resorting to the text.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

1. Introduction: In the past, scholars have utilized NDVI for extracting sugarcane. What specific problem did you address by using NDVI for sugarcane extraction?

2. Methods: In this study, the extracted area of sugarcane was compared with the official statistical data to assess the accuracy of sugarcane mapping. However, it should be noted that this comparison may not accurately reflect the differences between the mapped area and the actual area of sugarcane due to variations in spatial extent and the possibility of similar areas.

3. Results: It is important to note that the accuracy of the study's results cannot be solely determined based on linear regression analysis conducted on the mapping area of only seven regions compared to the surveyed areas. The limited number of validation regions undermines the ability to accurately assess the precision of the study's findings.

4. Discussion: The study proposes an NDVI-based sugarcane extraction method, NDSI, and compares its accuracy with RF and SVM. The results show that the proposed method outperforms the other two methods in terms of accuracy. However, it is important to note that the applicability of the proposed method is limited to local sugarcane extraction. The applicability of the proposed method for extracting sugarcane in other regions remains uncertain. I am unsure whether this method is universal and can be applied to extracting sugarcane in different regions. Given the broader applicability of RF and SVM methods, it is not possible to definitively conclude the superiority of the proposed method over RF and SVM. It can only be inferred that the proposed method may be more effective for local sugarcane extraction, based on the selected validation data used in the study.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

All required comments have been addressed by the authors, affirming that the article is now ready for publication.

 Moderate editing of the English language required