Synthetic Forest Stands and Point Clouds for Model Selection and Feature Space Comparison
, Isaac Nealey 2, Michael R. Gallagher 3, Nicholas S. Skowronski 4 and Brenden E. McNeil 1Round 1
Reviewer 1 Report
The challenges of field validation data and actual LiDAR data acquisition make it difficult to evaluate the best algorithm for characterizing forest stand volume using LiDAR. Therefore, this paper demonstrates the effects of using synthetic forests and simulated terrestrial laser scanning (TLS) to evaluate machine learning algorithms, scan configurations, and feature spaces that can best characterize forest stand volumes. It is concluded that the random forest and support vector machine algorithms generally outperform the k-nearest neighbor algorithm in estimating the square-level vegetation volume, regardless of the input feature space or the number of sweeps. Measurements that characterize occlusion using spherical voxels generally provide higher predictive performance than characterizing the vertical distribution of echoes using highly segmented summary statistics. It has a certain practical significance.
Enclosed are some suggestions for the paper:
It is recommended to list the algorithms to be compared in the article and briefly introduce these algorithms.
The techniques and tools used in the article, such as Blainder in 3.2, suggest that their functionality be briefly described in the text.
Some similar works have been published, such as “Simulation on Different Patterns of Mobile Laser Scanning with Extended Application on Solar Beam Illumination for Forest Plot” and “Simulation of multi-platform LiDAR for assessing total leaf area in Tree crowns”. These methods could be mentioned in the Introduction section. Meanwhile, more quantitative analysis could be added to develop an exploration of the best scanning patterns.
Tree species including conifer and broad-leaved trees could be brought into the experiments with different aspect assessments.
Some quantitative results could be articulated in the Conclusion section.
Author Response
"Please see the attachment."
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This manuscript demonstrates the use of synthetic forests stands and simulated terrestrial laser scanning (TLS) for the purpose of evaluating which machine learning algorithms, scanning configurations, and feature spaces can best characterize forest stand volume. The manuscript is a valuable article for contributing analysis on the study of estimating forest stand volume using TLS.
1. Only one of the keywords, TLS or terrestrial laser scanning, is sufficient.
2. Before concluding "As such, there is a need for accurate information for quantifying forest resources and monitoring their dynamics", only one reason is provided. Before concluding "As such, there is a need for accurate information for quantifying forest resources and monitoring their dynamics", only one reason is provided.
3. "Synthetic data" is mentioned for the first time in the article, and it is suggested that it be briefly explained.
4. In Section 3.1, the reasons for choosing mixed evergreen deciduous forests from a modeling perspective should be added.
5. The introduction section is an important overview of the background and significance of the article, rather than a simple list of other researchers' studies, and the authors need to improve the introduction section.
6. The software appearing in the manuscript should be added to the company and city information.
7. Pay attention to checking the capitalization of words in the manuscript.
Minor editing of English language is required in this manuscript.
Author Response
"Please see the attachment."
Author Response File: Author Response.pdf
Reviewer 3 Report
Dear Authors,
Thank you for submitting your manuscript entitled “Synthetic Forest Stands and Point Clouds for Model Selection and Feature Space Comparison” to Remote Sensing. I have reviewed your manuscript and found it to be generally well-written and informative.
However, there are some minor issues that could be addressed to improve the overall quality of the manuscript.
Firstly, it would be helpful to provide more detailed information on the methods used to generate the synthetic forest stands and point clouds.
Secondly, it would be beneficial to include more discussion on the limitations of the study (like in mountain forest) and potential avenues for future research.
Additionally, Figure 7 should clearly indicate the subfigure labels (a) and (b) directly.
Overall, I believe that this manuscript has the potential to make a valuable contribution to the field of remote sensing. I look forward to seeing the revised version of your manuscript.
Proofreading is needed.
Author Response
"Please see the attachment."
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
As a whole, I feel that the revisions and additions to the manuscript add value to both its readability and scientific merit. The authors did a nice job of synthetic forest stands and point clouds for model selection and feature space comparison. Overall, the elaboration is comprehensive as it pertains to sensing secondary metabolites. The manuscript is well written, and the revisions made to the current draft make it better suited for publication.
