Next Article in Journal
Photovoltaic Panel Parameter Estimation Enhancement Using a Modified Quasi-Opposition-Based Killer Whale Optimization Technique
Previous Article in Journal
Carbon Emission Reduction Effects of Government Talent Attraction Policies: Evidence from Fujian Province, China
Previous Article in Special Issue
Microbial Additive Isolated from Exotic Semi-Arid Cactus and Cottonseed Byproduct in Sustainable Sorghum Silage Production
 
 
Article
Peer-Review Record

Assessment of Organic Matter Content of Winter Wheat Inter-Row Topsoil Based on Airborne Hyperspectral Imaging

Sustainability 2025, 17(11), 5160; https://doi.org/10.3390/su17115160
by Jiachen He 1,2, Wei Ma 2,* and Jing He 3
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3:
Reviewer 4: Anonymous
Sustainability 2025, 17(11), 5160; https://doi.org/10.3390/su17115160
Submission received: 10 April 2025 / Revised: 17 May 2025 / Accepted: 29 May 2025 / Published: 4 June 2025

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper focuses on the demand for monitoring the distribution of soil organic matter and conducts research on soil organic matter monitoring methods based on hyperspectral imaging analysis. On the basis of partial sampling and detection of soil organic matter, hyperspectral image analysis and processing are carried out. Through image preprocessing, fitting analysis, and band selection optimization, suitable soil organic matter detection bands and parameters are determined, and the distribution of soil organic matter is obtained through model inversion. Based on the NDVI map and organic matter content map, the relationship between soil organic matter and crop aboveground biomass is jointly obtained. The research paper has certain promotion and application value.

There are some problems with the paper:

  1. In the fitting of the relationship between spectra and soil organic matter, the soil organic matter monitoring results (SOM) should provide more intuitive distribution data than Table 1, such as a distribution map, for comparison with subsequent spectral monitoring inversion.
  2. The fertilization plan for plot 1 should specify the types, quantities, distribution, and timing of fertilization.
  3. In the spectral image acquisition of wheat fields, it is only collected once during the wheat sprouting period. To get the relationship between soil organic matter distribution and aboveground biomass, the NDVI is important. It may be difficult to get good NDVI index and SOM at the same time with only once imaging. Why not to collect wheat images again when it grows to a certain height? It is better to provide some example images of wheat growth during collection, as well as spectral images of different time periods and bands for field No. 1 and field No. 2.
  4. The paper does not explain how spectral images correspond to sampled soil organic matter data in terms of preprocessing, band selection, and fitting analysis. Is the image based on the location of the sampling point, the neighborhood information near the sampling point, or all spectral information of the entire plot?
  5. Some images are not clear, such as Figure 2, Figure 3, etc. Figure 2 does not clearly indicate which is field No. 1 and which is field No. 2, and the latitude and longitude coordinates of the two plots cannot be seen clearly. The length, width, area, and other information of each field should be provided. The spectral information in Figure 3 does not specify which region in the field it belongs to.
  6. Figure 4 shows the horizontal and vertical axes representing the data that should be explained. Why are they all soms? How is the relationship between the spectrum in Figure 4 and som calculated? The corresponding relationship between the fitted data points and spectral points should be appropriately explained.
  7. In L11” s The spatial distri-”, the “s” is redundant.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Review comments.

  1. In the instruction, some studies on the application of hyperspectral remote sensing technology and machine learning methods in soil monitoring were listed. However, the summary and comparative analysis of these literatures were rather general. For instance, the similarities and differences in data acquisition methods, model construction methods, and application effects among different types of research were not clearly pointed out. This made it difficult to highlight the innovation and advantages of this research. It is suggested to conduct a more detailed sorting and comparison of the relevant literatures, emphasizing the unique aspects of this research in terms of research methods, technical routes, or application fields, so as to provide a more solid theoretical foundation for the subsequent research content. The authors should add the clear relevance of the study in the introduction.
  2. In the "2.3. Hyperspectral Data Acquisition" section, it is mentioned that data was collected using a UAV-mounted hyperspectral imager. However, the specific details regarding the flight speed, route planning, etc. were not provided. These factors may affect the quality and representativeness of the data collection. It is recommended to provide these details to make the experimental part more detailed and rigorous.
  3. Suggest to clarify statistical methods: The text mentions multiple times that data analysis and modeling are conducted. For instance, in "3.4. Regression Model Analysis", ridge regression and Lasso regression models are employed. However, the statistical test methods (such as significance tests, etc.) for these models have not been clearly specified. It is necessary to supplement the statistical methods used and their results to ensure the reliability and scientific nature of the research conclusions.
  4. In the "4. Discussion" section, although discussions were conducted on characteristic bands, model selection, and the interaction between soil and crops, they were somewhat brief. For instance, in "4.3. Analysis of the Impact of Other Factors", it was mentioned that the NPK and pH values of soil have an impact on SOM. It is suggested that a more detailed analysis be conducted on the differences in the influence of various factors on SOM under different fertilization conditions and the underlying mechanisms, and to further explore the interaction relationships among these factors, so as to make the discussion more profound and comprehensive.
  5. The authors lack a comparative discussion with other similar methods in the model selection section of the Discussion section, and suggest adding a comparison with related literature.
  6. The overall English description of the article needs to be improved. It is suggested to find an English native speaker to optimize the writing of the paper.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

This study establishes a rapid detection method for soil organic matter (SOM) content, reveals its positive correlation with wheat biomass, and provides support for precision agriculture and hyperspectral imaging applications. However, the following issues need further clarification from the authors:

 

  1. Introduction - You should begin with the problem, the gap, then propose the research question and just after that say what they want to do to address that. Where is the gap? And you should clearly why it is a gap? Once again, if you say that it is a gap, then try to build a case for the gap.
  1. In the introduction, please add the main contributions of this paper, further clarifying the research innovations and the key issues addressed.
  2. In Section 2.3, the 10 nm spectral interval (especially in the shortwave infrared region) may miss key organic matter-sensitive bands (e.g., in the visible to near-infrared region), and the initial spectral range (884-2442 nm) does not cover the visible light range (400-700 nm), while some characteristic bands of soil organic matter (e.g., humic acid absorption peaks) may lie within the visible range. It is recommended to extend the spectral range to 400-2442 nm, reduce the spectral interval (e.g., 2-4 nm), or verify whether the selected bands cover the organic matter-sensitive region.
  3. The authors only used Ridge regression and Lasso regression methods to estimate soil organic matter (SOM) content, lacking a comparative analysis with other advanced methods, such as using convolutional neural networks (CNN) for SOM estimation. The following reference may be useful:  

   â‘  Li H, Ju W, Song Y, et al. *Soil organic matter content prediction based on two-branch convolutional neural network combining image and spectral features* [J]. Computers and Electronics in Agriculture, 2024, 217: 108561.

 

  1. The authors used Ridge regression and Lasso regression to estimate soil organic matter (SOM) content, only analyzing the correlation between SOM content and wheat biomass and establishing a quantitative model for both. It is suggested that the authors include a related analysis and results on the estimation accuracy of SOM content to enhance the completeness and persuasiveness of the study.
Comments on the Quality of English Language

The overall quality of the English language in the manuscript is acceptable, and while there are some areas where language could be improved (e.g., grammar, sentence structure), these issues do not significantly affect the comprehension of the content. The core ideas and conclusions are still clearly communicated.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

The paper presents a study assessing organic matter content of winter wheat inter-row topsoil. The study was based on airborne hyperspectral imaging.

Introduction: The authors presented in detail the study’s background.

Materials and Methods: The authors presented the case study area and applied methods. This section is divided in subsection clearly demonstrating all methodological aspects.

Results and Discussion: The results are compelling and presented in detail with figures providing additional insights into the study findings.

Discussion: The discussion is based on the study results. The authors divided this section into 4 subsections tackling the key issues presented in the study. This approach made the reading easier.

Conclusions: The conclusions are presented in several points. In my opinion there should be added a paragraph on recommendations for farmers and policymakers as well as information on the study limitations and future study needs.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have thoroughly addressed all the concerns I previously raised and have made careful and detailed revisions to the manuscript. The quality of the paper has improved significantly; it is now well-structured, logically coherent, and rich in content. It meets the standards for publication in the journal. I recommend that the manuscript be accepted for publication.

Back to TopTop