Study on the Extraction of Topsoil-Loss Areas of Cultivated Land Based on Multi-Source Remote Sensing Data
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis study focuses on soil issues, particularly the impact of topsoil loss on agricultural production. Taking Heishan Farm in Heilongjiang as the research object, it employs remote sensing technology and machine learning methods to design four extraction schemes for topsoil loss, integrating multi-source data and developing and optimizing identification models based on random forest and support vector machine algorithms. The results show that Scheme 4, which integrates spectral, topographic, and spectral index features, achieves the best effect. The random forest model has higher classification accuracy, and after optimization with the particle swarm optimization (PSO) algorithm, the PSO-RF model has the highest extraction accuracy and effect. The research results provide valuable references for agricultural ecological protection and sustainable soil resource management. However, there are still some details to be addressed before the manuscript is published. The main concerns and detailed comments are as follows.
1.The background description in the abstract is suggested to be simplified. The mention of "multi-source data" without specifying the data sources should be clarified by listing the data types to enhance the clarity of the expression.
2.In the introduction, the description of soil heterogeneity can be simplified, highlighting its direct impact on the study. When reviewing previous studies, focus should be on the most relevant achievements to this study's methods or objectives, avoiding excessive details.
3.It is recommended to clearly summarize the innovations and specific work of this study at the end of the introduction to highlight the study's relevance and forward-looking nature.
4.In Section 3.1, it is suggested to unify the representation of OA values in the text and charts to ensure consistent information expression. Simplify the presentation of data; specific percentage data on model accuracy improvement can be detailed in tables or charts.
5.In Section 3.2, it is recommended to supplement the analysis on whether features with lower importance can enhance their predictive power in the model through interaction or combination with other features.
6.In Section 4.3, when discussing future research directions and subsequent work, it should be clearly explained how these directions will specifically address the limitations of this study.
Author Response
请参阅附件。
Author Response File: 
Author Response.docx
Reviewer 2 Report
Comments and Suggestions for AuthorsThe paper integrates PSO with RF and SVM to identify soil erosion areas based on remote sensing data. It demonstrates a significant amount of work and the research content has important social value and practical significance. However, there are still the following shortcomings:
1. The definition of "topsoil lose area" is not clear enough. The paper uses Google Earth images for sampling, but there are no corresponding images to visually represent the sampling points.
2. The paper selects 18 features for analysis. It is suggested to use a table to comprehensively list them.
3. Please provide a detailed introduction on how to use PSO to optimize the RF and SVM models, and also elaborate on the parameter settings and adjustment processes of the RF and SVM models.
4. Enrich the results of the paper appropriately, and add a discussion on the research contributions in the discussion section, explaining the general applicability of the research methods.
Author Response
请参阅附件。
Author Response File: Author Response.docx
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript presents a valuable contribution to the field by integrating multi-source remote sensing data for topsoil loss extraction. The methods employed are sound, and the results demonstrate the potential of combining spectral features with topographic data. However, there are several areas where the manuscript could benefit from improvements to enhance clarity, depth, and the overall rigor of the analysis. Firstly, the absence of a clearly defined hypothesis limits the depth and systematic nature of the discussion. A well-articulated hypothesis would guide the analysis of results and provide a clearer framework for evaluating the findings. This would strengthen the overall argument and offer more insight into the relationship between the methods used and the outcomes observed. Secondly, While the discussion addresses key findings, it tends to be somewhat generalized. A deeper, more systematic analysis of the results in the context of the proposed hypothesis would provide greater clarity and understanding. Additionally, comparisons with existing literature and methodologies would further contextualize the study's contributions. Thirdly, Certain sections, especially in the methodology and results, are written in long and complex sentences that could be broken down for clarity. Simplifying these sentences would make the manuscript more accessible to a broader audience and improve the overall readability. Overall, the study is promising, but addressing the above and following points will help clarify the narrative, deepen the analysis, and make the manuscript more impactful.
Abstract
The abstract of this manuscript provides a comprehensive overview of the study on topsoil loss extraction based on multi-source remote sensing data. The research topic is relevant, particularly in the context of agricultural sustainability and soil protection, and the study makes a valuable contribution to the use of remote sensing and machine learning for environmental monitoring. Some sentences in the abstract are overly long and complex. For example, the sentence “Traditional survey methods are limited by time-consuming and labor-intensive processes, high costs, and complex data processing, making it difficult to meet the needs of large-scale research and information processing” could be broken into two shorter sentences for better readability. In addition, the results section could benefit from clearer presentation of the key statistical outcomes. Specifically, the statement “The PSO-RF model achieved an overall accuracy of 0.97 and a Kappa coefficient of 0.94” would be stronger if accompanied by some context about the significance of these results.
Introduction
Line 42-44: The first sentence is quite long and covers multiple ideas. Breaking it into two sentences would improve readability.
Line 49-53: The transition from discussing soil's heterogeneity to soil degradation issues could be clearer.
Line 58-61: The sentence is long and covers multiple ideas. It lacks sentence breaks, which impacts readability.
Line 61-63: The phrase "At the same time" is redundant. "Dominated by hydraulic erosion" is abrupt and lacks a smooth transition to the previous idea.
Line 63-66: The sentence structure is not tight enough. The phrase "black soil water erosion is developed" is vague and unclear. The impact on the thinning of the soil layer is not clearly articulated.
Line 66-69: "Sloping cropland is the main cropland resource" is wordy and could be simplified. The term "irrational land reclamation" is unclear and needs further explanation or rephrasing.
Line 73-76: "Making the farmland in the region face a severe crisis" is awkward and could be phrased more smoothly. "Fertile topsoil is extensively eroded and lost" is repetitive and could be simplified.
Line 76-79: "Vividly refer to" is informal and should be replaced with more formal academic language.
Line 79-82: "Mainly Quaternary loess-like clay deposits" could be more concise. "Constituting a potential desertification sand source" is vague and could benefit from more precise language or additional explanation.
Line 82-84: "Land sanding" is an uncommon term and should be replaced with "land degradation" or "desertification" for clarity.
Line 86-89: The sentence is long and covers multiple ideas. It lacks sentence breaks, which impacts readability.
Line 94-96: The sentence is repetitive, particularly with the phrases "predict the potential for soil degradation rather than actual areas of soil degradation" and "soil and water conservation factors," which could be streamlined for clarity and conciseness.
Line 97-100: The phrase "the spatial prediction accuracy and timeliness of these models are challenging to validate" is cumbersome and could be simplified for better readability.
Line 102-106: The sentence is overly long and contains multiple ideas, making it difficult to follow. Breaking it into two or more sentences would enhance readability.
Line 113-117: The sentence "The RF algorithm, which uses a combination of geologic type, geomorphic type, elevation, and slope, outperforms the SVM and MLC algorithms in terms of accuracy" is somewhat verbose and could be simplified for clarity.
Line 139-148: The paragraph does not clearly present any hypotheses. Typically, a hypothesis should be included to provide a framework for the research design and to guide the study’s objectives. In addition, while the paragraph mentions developing a new extraction technique and providing high-precision identification, the goals are still somewhat broad. It would be helpful to make the objectives more specific, such as stating the anticipated improvements in precision, the novelty of the method, or how it compares to existing techniques.
Materials and Methods
Line 2-3: The sentence is quite long, consider breaking it into two sentences.
Line 5-7: The phrase "The climate features a windy spring, a brief hot summer, a rapid cooling in autumn, and a long, cold winter" can be more concise. "A long, cold winter" and "a rapid cooling in autumn" could be combined or clarified for better flow.
Line 13-14: "The annual accumulated temperature above 10°C is between 2000-2300 degree-days, with a frost-free period of 115 to 120 days" could be more concise. "Accumulated temperature" could be better stated as "annual temperature above 10°C."
Line 170-173: The sentence is long and includes several technical details, which could be overwhelming for some readers. It could be split for better readability.
Line 175-177: The phrase "The Sentinel-2 satellite data used in this study offer ground resolutions varying from 10 m to 60 m, depending on the wavelength band" could be simplified for clarity. The expression "depending on the wavelength band" can be clarified or removed if not critical.
Line 179-181: The sentence is quite complex, with multiple ideas strung together. Breaking it into smaller sentences would improve clarity.
Line 201-204: The sentence "In order to reduce the interference of other terrestrial features and spectral information of different crops, as well as to avoid the imbalance of sample categories" could be more concise. It’s a bit long and could be clarified by breaking it into separate parts.
Line 239-242: The sentence structure is quite dense, and listing several methods in one sentence makes it harder to follow. Breaking the methods into smaller parts or using bullet points could improve the readability.
Line 250-253: The sentence "It can be observed that there are certain differences in the spectral characteristics between the surface-degraded soil and the normal soil, but the overall trend is consistent" is a bit wordy and repetitive. "Certain differences" and "overall trend" can be simplified.
Line 294-296: The sentence "In multiple experiments, the performance of EVI, RVI, MSI, and BSI did not meet expectations, hence these indices were eliminated" is somewhat casual and could be reworded to sound more formal and academic.
Line 340-359: If the mathematical formulas are not original to the author and are instead drawn from other sources, it is generally more appropriate to cite the reference rather than including the formula itself in the methodology section.
Line 361-378: same as the above comment.
Results
Line 400-402: The phrase "performs poorly in identifying TLA under complex surface conditions" could be more specific. It would be helpful to elaborate on what is meant by "complex surface conditions" or what issues are being encountered.
Line 402-405: The sentence "In contrast, Schemes 2 and 3 improve the identification range and accuracy by incorporating topographic features and related indices" is somewhat vague. It could be clearer about how topographic features and indices improve performance and should explicitly mention what "identification range" refers to.
Line 444-447: The sentence "Following closely are bands 4, 3, 6, and 8, which cover the visible to near-infrared spectrum and are essential for capturing soil, vegetation conditions, and surface structure information closely related to topsoil loss" is quite complex. Breaking it into two sentences could improve readability and comprehension.
Line 447-449: The phrase "Additionally, elevation, as a topographic factor, shows its importance weight and indicates that topographic changes are closely related to water erosion and sedimentation processes" could be reworded for clarity. "Importance weight" might be simplified or clarified.
Line 449-451: The explanation for the lower importance of slope, relief, and roughness could be expanded to clarify why these features are less significant, and the explanation of their relationship with elevation data should be more direct.
Discussions
Line 512-515: The explanation of terrain slope's impact on soil erosion and water loss is a bit complex. It could be broken down into two sentences to better articulate the relationship between slope, terrain, and TLA formation.
Line 556-559: The sentence "These features could potentially assist in delineating the spatial heterogeneity within agricultural fields" is somewhat vague. It could benefit from more specific examples or context, explaining what spatial heterogeneity is and how the features assist in delineating it.
While the discussion provides valuable insights into the performance and potential of the models used, it lacks sufficient depth and systematic analysis, particularly in the absence of a clearly defined hypothesis. A well-formulated hypothesis would have provided a more focused framework for the discussion, guiding the analysis of results in a more structured manner.
Author Response
请参阅附件。
Author Response File: Author Response.docx
Round 2
Reviewer 3 Report
Comments and Suggestions for AuthorsI am satisfied with the revision
