Multi-Scale Assessments and Future Projections of Drought Vulnerability of Social–Ecological Systems: A Case Study from the Three-River Headwaters Region of the Tibetan Plateau

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I enjoyed reading the paper about drought in the Three-River Headwaters Region. The use of the SES framework and the three dimensions of exposure, sensitivity, and adaptive capacity is appropriate. The focus on the village, town, and county levels gives granularity and a rich analysis. The use of methods is appropriate but could be made clearer. With that said, here's a few thoughts:

• Abstract can be made more concise.
• Line 127 needs to be reworded.
• Why indicators fall within a certain domain can be better justified. For example, why are cropland, grassland, and wetland areas considered measures of exposure, while per capita area measures of the same land type considered measures of adaptive capacity?
• The methods section should be more explicit in saying at the beginning what the unit of analysis is. I would therefore place section 2.4 Data Sources at the beginning of the Methods section.
• The methods section needs to be more explicit on what problem the listed analytic methods attempt to solve (entropy weight method). Why do we need to assign indicators different weights? Why not assign all indicators subjective weights? What does "robustness" of an indicator in this context represent? Seems here that robustness is referring to less variance in the weight as the output for the models, but it's not entirely clear to me why this is the most important factor.
• It's not clear what the y-axis is Figure 2 represents. Is it the ranking of each county on the index?
• It is not clear what stakeholder input was used in the MCE model, nor how it was collected or used. This should be clarified in the methods or results section.
• Discussion should include a bit more relevance to policy. What should we make of this improvement in drought over time the region and the specific clustering pattern? How should we interpret the importance of specific variables to the vulnerability indicators? Are those effects causal and does it mean we need to improve these variables to lower drought vulnerability? 
• I would add a short limitations section to describe ways the validity of the findings could be threatened. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Review Comments to the Authors

Dear Authors,

Thank you very much for submitting your manuscript titled “Multi-scale assessments of drought vulnerability of Social-Eco-logical Systems: A case study from the Three-River Headwaters Region of the Tibetan Plateau.” After carefully reading and evaluating your work, I believe that your study holds significant scientific and practical value, especially in providing new perspectives and methods for assessing drought vulnerability in socio-ecological systems. Below are my specific review comments:

Strengths:

1. Comprehensive Selection of Evaluation Indicators: You have selected a wide range of evaluation indicators that cover social, ecological, and economic dimensions. This multidimensional indicator system can comprehensively reflect the characteristics of drought vulnerability in the Three-River Headwaters Region, providing a solid foundation for your study.
2. Appropriate Use of the Entropy Weight Method: By employing the entropy weight method to adjust the influence of different indicators on the results, you have effectively avoided biases associated with subjective weighting. This approach enhances the objectivity and scientific nature of your results. Your description and calculation process of the entropy weight method are also clear and reasonable.
3. In-depth Analysis of Spatial and Temporal Variations: You have not only analyzed the spatial distribution of drought vulnerability but also examined its temporal changes. This integrated approach of spatial and temporal analysis can more comprehensively reveal the dynamic patterns of drought vulnerability, providing important references for regional water resource management and ecological conservation.

Weaknesses:

1. Disorganized Reference Citations: During my review, I noticed some issues with the reference section. The organization of some references is not clear, which may affect readers' understanding of the research background and related work. I suggest that you unify the format of your references according to internationally accepted citation styles and categorize them by theme or chronological order to improve readability.

Suggestions:

1. Reorganize and standardize your references to ensure consistency in citation format. Additionally, provide more detailed interpretations of key references in the text to highlight their relevance to your study.
2. In the discussion section, you may further compare your results with those of similar studies to highlight the innovation and unique contributions of your research.

Overall, I believe your manuscript has high academic value and only requires minor revisions to the reference section to meet the publication standards. I look forward to seeing your careful consideration and improvement of the above comments.

Best regards,

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This paper focuses on the drought vulnerability of the socio-ecological system in the Sanjiangyuan region of the Tibetan Plateau, with clear logic and reasonable methodology, but there is still room for improvement in terms of index system, data, model and research depth.

Dimensionality of indicator system construction: In the sensitivity dimension, the selected indicators mostly focus on land use and industrial structure, and the internal structure and function of the ecosystem are not considered enough. For example, species diversity indicators are not included, while in arid environments, ecosystems with high species diversity are more stable and less sensitive to drought.

Representativeness of indicators: In the dimension of adaptive capacity, the traffic index only selects the mileage of highways, which cannot fully reflect the adaptability of regional transportation to drought response. Transportation modes such as air transport and rail transport also play an important role in key links in the response to drought, such as the deployment of materials and the movement of people, but they are not included.

Indicator overlap: the number of livestock in the exposure dimension and the proportion of large livestock in the sensitivity dimension, and there is some overlap between the two. The number of livestock affects the exposure, and the proportion of large livestock affects the sensitivity, which may lead to double counting of the degree of influence of related factors and interfere with the analysis results.
In terms of data, data update lags: data mainly rely on statistical yearbooks, government gazettes, etc., and their update cycles are usually long. For example, some meteorological data and socio-economic data may not reflect the rapid climate change and socio-economic restructuring in recent years, which will affect the accurate assessment of the current drought vulnerability.

Differences in data accuracy: The accuracy of data varies from different data sources. The eco-meteorological bulletin data may be based on monitoring at a larger spatial scale, while the socio-economic data in the government bulletin may be refined to the county and town level, and this difference in accuracy may introduce errors in the process of data integration and analysis.

Simple handling of missing data: The processing method of missing data values is not detailed in this article. If a large amount of data is missing, simple filling or deletion operations may change the distribution characteristics of the data and affect the reliability of the analysis results.

Limitations of model application model assumptions: The entropy-weight method assumes that the indicators are independent of each other, but in reality, there are complex associations between the factors of the socio-ecological system. For example, agricultural and livestock outputs are closely related to land use types and water resources, which may lead to inaccurate weights determined by the entropy weight method.

Single Forecast Model Scenario: When the MCE-CA-Markov model predicts drought vulnerability in 2050, the scenario settings only consider historical trends and stakeholder-based adaptive governance rules. Extreme climate change scenarios, such as persistent high temperatures and droughts, abrupt changes in precipitation patterns, and the impact of technological breakthroughs (such as the spread of new irrigation technologies) on drought vulnerability are not considered.

Insufficient model validation: After using multiple models for analysis and prediction, the validation of model results is insufficient. The drought vulnerability index is calculated only by the entropy weight method selected by robustness analysis, and it is difficult to determine the reliability of the model results due to the lack of comparison and verification with actual observation data or other reliable research results.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf