Estimation of Leaf Area Index for Dendrocalamus giganteus Based on Multi-Source Remote Sensing Data

Round 1

Reviewer 1 Report

Suggestions and comments on the article can be found in the attached file.

Comments for author File: Comments.pdf

The text needs to be revised in terms of English, some sentence and paragraph formulations need to be revised. The punctuation in some passages also needs to be improved, as well as the terms used in the field.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors employed machine learning techniques, such as Random Forest (RF), Gradient Boosting Regression Trees (GBRT), and K-nearest neighbors (KNN), to estimate Leaf Area Index (LAI) for mapping Dendrocalamus giganteus using remote sensing data (i.e., ICESat-2/ATLAS, Sentinel-1, and Sentinel-2). The topic is compelling; however, the novelties should be highlighted, and recent papers should be cited. Here are my comments, which should be addressed:

1. Examples of Remote Sensing Data for Forests and Data Fusion:
• Please include examples of using remote sensing data for forest applications and explain the importance of data fusion. For instance, combining Synthetic Aperture Radar (SAR) and optical images can enhance forest mapping accuracy by leveraging the complementary information provided by each data type. SAR provides structural information through cloud-penetrating capabilities, while optical images offer detailed spectral information. For further insights, refer to recent studies, such as (https://doi.org/10.1016/j.kjs.2023.11.008) and ) https://doi.org/10.1007/s11119-022-09893-4), which discuss the benefits of using SAR and optical remote sensing images in forest applications.
2. Literature Gaps and Study Justification:
• Clearly delineate the literature gaps and justify the necessity of this study. Please use this work (https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018RG000608). Highlighting the research gaps will underline the significance of your work. List the novelties or objectives of your research study, ensuring that each aim addresses a specific question or gap. For instance, if the current literature lacks a comprehensive approach to integrating ICESat-2/ATLAS with Sentinel-1 and Sentinel-2 for LAI estimation, emphasize this gap. Ensure that the experimental results explicitly address these aims.
3. Concise Study Area Description:
• The study area description is currently too lengthy. Please condense this section to focus on the essential details relevant to your study.
4. Remote Sensing Dataset Description:
• The explanation of the remote sensing dataset is overly detailed. Reduce this section and incorporate relevant citations to streamline the information.
5. Flowchart and Terminology:
• Use "spectral bands" instead of "band factor" in the flowchart and explanations. Clarify the term "correction analysis"; if you mean preprocessing, specify this clearly. Revise the flowchart to ensure it accurately represents the estimation strategy. A well-designed flowchart is crucial for conveying your methodological approach effectively.
6. Model Accuracy Evaluation:
• To evaluate model accuracy, include the Relative Root Mean Square Error (RRMSE) by citing this study (https://doi.org/10.1016/j.kjs.2023.11.008). Present the results based on this metric to provide a more comprehensive model performance evaluation.
7. Feature Selection Method:
• Incorporate a feature selection method. Using an excessive number of features can negatively impact estimation performance. Explain your feature selection process and its importance in enhancing model efficiency and accuracy.
8. Discussion on R² and Data Distribution:
• In Table 5, the R² values exceeding 0.8 and 0.9 do not imply a perfect correlation between field and remote sensing data. Reflect on this in your discussion, addressing the potential issues of obtaining test and train data. Many studies report high R² values, which may not reflect real-world performance. Discuss the distribution of train and test data, possibly including a map visualization to illustrate this distribution and enhance the transparency of your methodology.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors responded satisfactorily to the suggestions made. 

Reviewer 2 Report

The authors well addressed my comments.