Round 1

Reviewer 1 Report

The research presents a real-time (AI) model to evaluate the landslide hazard due to rainfall in the site of photovoltaic power plants using combined technique of ANP-FBN-AHP. the authors have done a great job merging these concepts together in one technique and the output seems reasonable and makes perfect sense. however, there are some minor comments to be considered before the acceptance as follows:

1- What is "DEM" in line 138 ?

2- Why do you considered the minimum soil thickness for C5 not the maximum which is critical or even the average ?

3- Also why the average rainfall for C7, the maximum is more critical ?

4- Where did you get the weights in Table 2?

5- You mentioned an "domain experts in landslides" in line 272, this should be described in details, how many expert, their qualifications, areas of expertise, affiliations, conflict of interest, how did they share their opinions (open questionnaire, structured questionnaire,  Delphi method, ..) and a sample from the  questionnaire should be added in the appendix.

6- Where did you get the values in Table 3? give reference or verification

7- in Eq. 5, is the critical state at probability of landslide of P=0%  or  P=80% ?

8- In the introduction section, some related recent references are missing such as: https://doi.org/10.1080/23311916.2022.2122158
http://dx.doi.org/10.28991/CEJ-2023-09-02-015
https://doi.org/10.1155/2021/3267018
https://doi.org/10.3390/su142416740

9- the research is mainly concerned in a certain particular site, hence it is a "Case study". this should be mentioned in the title and the abstract 

10- in the conclusion section, it should be clearly mentioned that the outcomes of this research are valid only for the considered case study, and otherwise they should be verified.

11- Also, conclusion section should include the current study limitations and recommendations for further studies.   

Author Response

Response to Reviewer Comments

 

Point 1: What is "DEM" in line 138 ?

Response 1:  The author has made a supplementary explanation of "DEM" in the manuscript. The meaning of "DEM" is Digital Elevation Model. It is a 3D model that contains elevation data, representing the height or elevation of the Earth's surface at various points on a grid. It is a fundamental dataset used in Geographic Information Systems (GIS) for analyzing and modeling terrain characteristics, such as slope, aspect, and drainage patterns. (see Line 149)

 

Point 2: Why do you considered the minimum soil thickness for C5 not the maximum which is critical or even the average ?

Response 2:  The consideration of the minimum soil thickness for C5 in our study for the following reasons. The focus of our research was to assess the landslide susceptibility of large photovoltaic power plants under extreme rainfall conditions. In this context, we aimed to identify the areas with the highest susceptibility to landslides, which are typically associated with thinner soil layers. The minimum soil thickness is crucial in landslide susceptibility assessment because it represents areas where the underlying bedrock is more exposed, leading to increased instability and a higher probability of landslide occurrence.

 

Point 3: Also why the average rainfall for C7, the maximum is more critical ?

Response 3:  We appreciate the feedback from the reviewers. Due to the location of the study area in a high altitude zone, the number of rainfall monitoring stations set up in the region is very sparse. Therefore, it is difficult to obtain the maximum rainfall in the study area. Meanwhile, on the basis of previous studies, the authors finally selected the representative average rainfall from the historical rainfall data (2002-2022) as one of the evaluation factors of landslide susceptibility.

 

Point 4: Where did you get the weights in Table 2?

Response 4:  The authors used the ANP （Analysis Network Process） method to calculate the indicator weights of the criterion and network layers in the vulnerability evaluation indicator system based on the landslide vulnerability evaluation indicator system. By multiplying these weights, the comprehensive weights were derived, as presented in Table 2.

 

Point 5: You mentioned an "domain experts in landslides" in line 272, this should be described in details, how many expert, their qualifications, areas of expertise, affiliations, conflict of interest, how did they share their opinions (open questionnaire, structured questionnaire,  Delphi method, ..) and a sample from the  questionnaire should be added in the appendix.

Response 5:  We appreciate the reviewers' valuable feedback and understand the importance of providing detailed information about the domain experts involved in our study. In response, we have updated our methodology section to include comprehensive descriptions of the experts and their contributions to the research，and the link to the e-questionnaire was also added. (see Line 283-288)

 

Point 6: Where did you get the values in Table 3? give reference or verification

Response 6:  Table 3 shows the assignment table of a parameter (slope coefficient k) in the power generation loss calculation formula (Equation 3), which is obtained by the author based on the geographical location and sunshine law of the study area. The research area located in the Eastern Hemisphere and the Northern Hemisphere. Combined with the sunshine law, it can be seen that the sunshine of the southern slope in the research area is more abundant than that of the Northern Slope. Therefore, when the slope direction is due south, the value of k is the largest, and gradually decreases as the slope direction changes to the north.

 

Point 7: in Eq. 5, is the critical state at probability of landslide of P=0%  or  P=80% ?

Response 7:  According to the definition of the rainfall intensity-rainfall duration (I-D) curve, a landslide is considered likely to occur when the probability of landslide P≥0%, i.e., P = 0% is the critical state.

 

Point 8: In the introduction section, some related recent references are missing such as: 

https://doi.org/10.1080/23311916.2022.2122158
http://dx.doi.org/10.28991/CEJ-2023-09-02-015
https://doi.org/10.1155/2021/3267018
https://doi.org/10.3390/su142416740

Response 8:  We thank the experts for listing the references, which has benefited the authors by reading them and using them as references in the introduction to enrich and elaborate on the application and current status of each research method. (see Line 518, 520 and 574)

 

Point 9: the research is mainly concerned in a certain particular site, hence it is a "Case study". this should be mentioned in the title and the abstract 

Response 9:  In this study, although the established evaluation index system has some specificity, the whole paper mainly discusses the method and process of disaster dynamic assessment. Meanwhile, the validation results show that the method proposed in this paper has good accuracy and robustness, and can be used as a general method in the field of disaster dynamic assessment.

 

Point 10: in the conclusion section, it should be clearly mentioned that the outcomes of this research are valid only for the considered case study, and otherwise they should be verified.

Response 10:  According to the expert opinions, the author modified the conclusion to some extent, and clearly stating that in this study the object is a photovoltaic power plant. (see Line 454-459)

 

Point 11: Also, conclusion section should include the current study limitations and recommendations for further studies.

Response 11:  Thank you for the valuable feedback from the reviewers. According to the expert opinions, the author modified the conclusion to some extent. Meanwhile, the author pointed out that in this study, due to the subjectivity of AHP method, the assessment results of the vulnerability of photovoltaic sites are susceptible to subjective errors, and it is insufficient to only use the rainfall intensity - duration (I-D) threshold as the criterion for the probability of landslide occurrence. Subsequently, the author intends to use a variety of evaluation methods and rainfall threshold curve to carry out landslide dynamic risk assessment zoning, and to select the most accurate and reasonable evaluation method by comparing the evaluation results. (see Conclusion)

 

We sincerely thank the reviewers for their insightful comments and believe that these revisions significantly strengthen the validity and reliability of our research findings.

Author Response File: Author Response.pdf

Reviewer 2 Report

 

The manuscript conducted a dynamic landslide risk assessment at a large photovoltaic power plant project under extreme rainfall conditions. It is a very interesting topic. Several comments are given to improve the study.

 L45. It is not common to talk about “the study area” when it is not introduced.  

 L 46 The impacts of intensive human activities (such as urban expansion) on the landslide susceptibility, and their exposure to landslide risks should be highlighted in the introduction parts. Following literatures are suggested.

      Evaluating trends, profits, and risks of global cities in recent urban expansion for advancing sustainable development. Habitat International……

Integration of landslide hazard into urban planning across Europe……

 The right panel of Fig.1 is unclear and uninformative.

 L219  Selection of “landslide susceptibility” assessment factors should be “landslide hazard vulnerability”

 Further discussion is needed to tell the advantages and limitation of the proposed methods,

 Comparative tests using other methods are suggested to show the merits of the proposed method.

Lastly, the construction of photovoltaic power plant had any impacts on landslide susceptibility? I mean you’d better tell more characteristics of the study related to photovoltaic power plant.

 

 

English is very difficult to understand/incomprehensible.

Author Response

Response to Reviewer Comments

Point 1:  L45. It is not common to talk about “the study area” when it is not introduced.

Response 1: Thank you for the valuable feedback from the reviewers. The authors have revised the introduction accordingly to avoid discussing the study area without going into detail. (see Line 41-47)

 

Point 2: L 46 The impacts of intensive human activities (such as urban expansion) on the landslide susceptibility, and their exposure to landslide risks should be highlighted in the introduction parts. Following literatures are suggested.

      Evaluating trends, profits, and risks of global cities in recent urban expansion for advancing sustainable development. Habitat International……

Integration of landslide hazard into urban planning across Europe……

Response 2: Thanks for the referees' recommendation of relevant literature, from which the author benefited a lot, and took this as a reference to carry out the discussion and explanation of the impact of human engineering on the risk of natural disasters. (see Line 576 and 578)

 

Point 3: The right panel of Fig.1 is unclear and uninformative.

Response 3: Thanks to the reviewer's suggestions, the author has modified the original surface remote sensing image into Digital Elevation Model (DEM) data. (see Fig.1)

 

Point 4:  L219  Selection of “landslide susceptibility” assessment factors should be “landslide hazard vulnerability”

Response 4: Done. (see Line 229)

 

Point 5: Further discussion is needed to tell the advantages and limitation of the proposed methods,

 Response 5: Expert advice is greatly appreciated. According to the expert opinions, the author modified the conclusion to some extent. Meanwhile, the author pointed out that in this study, due to the subjectivity of the AHP method, the assessment results of the vulnerability of PV sites are easily affected by subjective errors, and there are shortcomings such as using only the rainfall intensity-duration threshold (I-D) as a discriminating criterion for the probability of landslide occurrence. Subsequently, the authors intend to use multiple evaluation methods with rainfall threshold curves to carry out the dynamic risk assessment of landslides, and prefer the most accurate and reasonable evaluation methods by comparing the evaluation results. (see Conclusion)

 

Point 6: Comparative tests using other methods are suggested to show the merits of the proposed method.

Response 6: Thank you for the valuable feedback from the reviewers. Methods such as ANP, FBN, and ANP used in this study have been widely used by a large number of researchers in the field of geo-hazard risk assessment, and their applicability has been well recognized by researchers. In addition, in the section of Results and Discussion, the authors provide a full comparative validation of the results of the landslide susceptibility assessment in the study area by means of representative validation parameters (ACC and AUC).

It has to be admitted that there is a lack of real geo-hazard and material damage data as a validation set for risk assessment, as the study object is in the construction process. However, this research work can still be used as an important part of the disaster prevention and mitigation work of photovoltaic power plants. Through the analysis and discussion of the susceptibility evaluation zoning and risk zoning under different rainfall conditions, it can provide theoretical support for the prevention of rainfall-type landslide outbreaks and the reduction of losses caused by the disaster.

 

Point 7: Lastly, the construction of photovoltaic power plant had any impacts on landslide susceptibility? I mean you’d better tell more characteristics of the study related to photovoltaic power plant.

Response 7: Thank you for the valuable feedback from the reviewers. We agree that it is important to provide more information about the construction of the photovoltaic power plant and its potential impacts on landslide susceptibility in the study area. In the construction process of photovoltaic power station, a large number of driven steel piles was used as the supporting mechanism of photovoltaic modules, so the construction of photovoltaic power station has less impact on the original terrain and natural environment. In the process of establishing the landslide susceptibility evaluation model of photovoltaic site, the factor of human engineering activities considered is land use, which can be classified by the surface cover (such as photovoltaic site, residential land, forest land and ancillary facilities, etc.) to realize the quantification of the impact degree of human engineering activities.

 

We sincerely thank the reviewers for their insightful comments and believe that these revisions significantly strengthen the validity and reliability of our research findings.

Author Response File: Author Response.pdf

Reviewer 3 Report

I can say that the authors did a great effort with their research and the paper has some valuable findings and contributions. It is suitable to be published under this journal if the below addressed comments be made:

1. More recent literature review is required to keep it up to date. 

2. A room for discussion of the obtained findings must be established.  

3. be more specific in your final conclusion.

4. The contribution must clearly be highlighted. 

In general, it is acceptable but more polishing is encouraging. 

Author Response

Response to Reviewer Comments

Point 1: More recent literature review is required to keep it up to date.

Response 1: First of all, we are very grateful for the approval of our research. The authors have replaced some outdated references with more recent research. (see Introduction)

 

Point 2: A room for discussion of the obtained findings must be established.

Response 2: Thank you for the valuable feedback from the reviewers. We have revised the original manuscript to refine the completeness of the analysis and discussion of the results obtained in this study.

In the discussion, in addition to the original analysis of landslide susceptibility, vulnerability and dynamic risk results, we have added some targeted disaster prevention and mitigation recommendations, with a view to providing certain technical guidance for the construction and stable operation of photovoltaic power plants. The main elements are as follows: (1) Focused monitoring is required for the central valley and eastern steep slope areas of the photovoltaic power plants, which are "very high" risk areas, and disaster risk inspections of these areas are conducted several times during the rainfall period, with a view to carrying out appropriate de-risking treatments before a disaster occurs. (2) Where conditions permit, emergency material storage nodes should be set up in advance in the study area, and research on emergency material storage management should be carried out to set up optimal paths for the transfer of personnel and the transportation of materials. The shortest possible time should be used to transfer injured personnel and repair damaged facilities after a disaster occurs, and emergency plans should be used to minimize losses. (3) The photovoltaic power plants construction department should cooperate with the disaster management department to establish a number of disaster monitoring stations in disaster-prone areas, and formulate targeted disaster prevention and management plans according to the specific conditions of the disaster-bearing body. Improve the protection programs for photovoltaic power modules, ancillary buildings, roads and other facilities within the site to improve disaster prevention capabilities and reduce the vulnerability of disaster-bearing bodies. (see Line 427-445)

 

Point 3: be more specific in your final conclusion.

Response 3: We are very grateful for the reviewer' comments. Based on the reviewer comments, the authors have made some modifications to the conclusions, which are mainly (a) illustrating the accuracy and robustness of the combined model based on the validation parameters; and (b) utilizing the area change data of each risk level to support the dynamic assessment of landslide risk during rainfall. (see Conclusion)

 

Point 4: The contribution must clearly be highlighted.

Response 4: Thank you for the valuable feedback from the reviewers. We understand the importance of highlighting the contributions of our research. We have revised the manuscript to provide a clear and concise description of the key contributions made in this study.

In the introduction section, we now include a dedicated subsection that explicitly outlines the contributions of our research. The main elements are as follows: (1) Employing a combination of statistical models and machine learning models to conduct landslide hazard vulnerability zoning in the study area. (2) Developing a vulnerability assessment index specifically tailored for landslide hazards in large photovoltaic power plant power plants and determining the probability value of landslide vulnerability. (3) Utilizing the empirical rainfall threshold model of the study area, incorporating daily updated rainfall data as dynamic indicators, and constructing a dynamic risk assessment model for landslide hazards in the study area.

Through the above research, we hope to provide certain theoretical support for the construction and stable operation of photovoltaic sites. At the same time, we can also provide methodological reference for the disaster prevention and management of similar large-scale engineering sites. (see Line 91-102)

 

We sincerely thank the reviewers for their insightful comments and believe that these revisions significantly strengthen the validity and reliability of our research findings.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The manuscript has been well revised.

Moderate editing of English language required