Spatial Network Heterogeneity of Land Use Carbon Emissions and Ecosystem Services in Chang-Zhu-Tan Urban Agglomeration

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript uses the Chang-Zhu-Tan urban agglomeration as a case study to present a comprehensive analysis of the spatial network heterogeneity between land use carbon emissions (LUCE) and ecosystem services (ES), utilizing advanced methodologies including an improved gravity model, social network analysis, and a random forest–driven mechanism evaluation framework. While the study features a well-structured research design, several key revisions are required to further enhance its academic contribution.
1. What’s the novel features of this paper? It mentioned three main contributions, but I could not agree. As it mentioned in the Section Introduction, there have been many articles studied these topics. What’s the relationship between these three key scientific questions and three research objectives.
2. In this study, CE from cultivated land, forest land, grassland, water bodies, and unused land are calculated using the direct CE coefficient method. However, the sources and temporal consistency of coefficients (Table 2) need justification.
3 A comparative study was not included enough. Hence, there appears to be little basis for concluding that the proposed method is more reliable than other methods, such as gravity model parameters, random forest implementation. In addition, How do the findings compare with other urban agglomerations? A brief comparative discussion would highlight the generalizability or uniqueness of the results.
4. The paper includes explanations of influencing factors, but the analysis of mechanisms remains insufficient, with results presented in a largely descriptive manner. It should enhance the discussion of causal mechanisms to deepen the theoretical contribution
5. Some grammatical and punctuation errors were found in the manuscript. Authors are therefore urged to have the entire manuscript reviewed and correct by a native English speaker.

Author Response

Comments 1: What’s the novel features of this paper? It mentioned three main contributions, but I could not agree. As it mentioned in the Section Introduction, there have been many articles studied these topics. What’s the relationship between these three key scientific questions and three research objectives.

Response 1: We sincerely thank the reviewer for their valuable feedback and suggestions. In response to the comment regarding the novelty of our paper and the relationship between the scientific questions and research objectives, we have made several important revisions. We have clarified the unique contributions of our study by emphasizing the integration of LUCE and ES networks within a unified framework, alongside the development of new heterogeneity indices to analyze the spatial relationship between LUCE and ES. This distinction highlights the originality of our work compared to existing studies. Additionally, we have strengthened the connection between the three key scientific questions and the research objectives. Each objective is now explicitly aligned with its corresponding question, ensuring a clearer logical structure: Objective 1 addresses Question 1, Objective 2 addresses Question 2, and Objective 3 addresses Question 3. These changes improve the clarity and coherence of the manuscript.

Comments 2: In this study, CE from cultivated land, forest land, grassland, water bodies, and unused land are calculated using the direct CE coefficient method. However, the sources and temporal consistency of coefficients (Table 2) need justification.

Response 2: We sincerely thank the reviewer for their insightful comment regarding the carbon emission (CE) coefficients. In response, we have clarified that the CE coefficients used in our study are derived from the IPCC 2006 Guidelines for National Greenhouse Gas Inventories, Volume 4: Agriculture, Forestry and Other Land Use (AFOLU), which provides standardized coefficients widely used in global carbon emission assessments. Additionally, we have explained that these coefficients have been consistently applied in regional carbon emission studies over the past decade, ensuring their temporal consistency and comparability across different time periods (2010–2023). We believe these revisions address the reviewer’s concerns and enhance the clarity and scientific rigor of the manuscript.

Comments 3: A comparative study was not included enough. Hence, there appears to be little basis for concluding that the proposed method is more reliable than other methods, such as gravity model parameters, random forest implementation. In addition, How do the findings compare with other urban agglomerations? A brief comparative discussion would highlight the generalizability or uniqueness of the results.

Response 3: We would like to thank the reviewer for their insightful feedback. In response to the comments regarding method comparison and regional comparison, we have made significant revisions to enhance the manuscript. We expanded the method comparison section by providing a more detailed discussion on why we chose our proposed method over others like the gravity model and random forest, highlighting their strengths and limitations. We explained how the gravity model fails to capture dynamic changes over time and the challenges of the random forest in dealing with spatial relationships, justifying why our approach is better suited for analyzing LUCE-ES network interactions. Regarding the regional comparison, we have enriched the discussion by comparing our findings with the Beijing-Tianjin-Hebei region based on studies by Li et al. (2022) and Guo et al. (2025). We discussed the similarities and differences in LUCE-ES network coupling between the regions, emphasizing the impact of local urbanization patterns, ecological restoration efforts, and socio-economic conditions. This comparison highlights the importance of region-specific ecological strategies and tailored policies, contributing to a more comprehensive understanding of LUCE-ES network dynamics. We believe these revisions significantly improve the manuscript by providing a deeper and more logical analysis of the methods and regional dynamics.

Comments 4: The paper includes explanations of influencing factors, but the analysis of mechanisms remains insufficient, with results presented in a largely descriptive manner. It should enhance the discussion of causal mechanisms to deepen the theoretical contribution.

Response 4: We sincerely thank the reviewer for their insightful feedback on our manuscript. In response to the suggestion regarding the analysis of causal mechanisms, we have expanded Section 4.5.2. Driving Mechanism Assessment to provide a more comprehensive discussion of the causal relationships between LUCE and ES networks. Specifically, we have introduced the concept of a feedback loop between urbanization and ecological restoration. As urbanization increases, it not only drives carbon emissions and land-use changes but also creates a growing demand for ecosystem services, triggering ecological restoration efforts. These restoration efforts, such as enhancing forest cover and carbon sequestration capacity, help mitigate the negative impacts of urbanization on ecosystem services. Over time, the effectiveness of these restoration efforts strengthens the resilience of the ES network, improving its spatial distribution and connectivity. This feedback loop shapes the structure and dynamics of LUCE and ES networks, determining the intensity of their interactions. We believe these revisions address the reviewer’s concerns and enhance the theoretical contribution of the paper by offering a more in-depth understanding of the mechanisms driving the interactions between LUCE and ES networks.

Comments 5: Some grammatical and punctuation errors were found in the manuscript. Authors are therefore urged to have the entire manuscript reviewed and correct by a native English speaker.

Response 5:  We would like to thank the reviewer for pointing out the grammatical and punctuation errors in the manuscript. We have carefully reviewed the entire manuscript and made the necessary corrections to ensure clarity and accuracy. Additionally, we have sought assistance from a native English speaker to further improve the language quality. We believe these revisions have significantly enhanced the readability and overall quality of the manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

The study presents a valuable and novel framework and is technically sound. However, several sections such as the method's justification, result interpretation, and policy operationalization would require clearer articulation, more robust sensitivity/uncertainty analysis, and linguistic polishing. Please refer to the attached file for my specific comments. 

Comments for author File: Comments.pdf

Author Response

We would like to express our sincere gratitude to the reviewer for your detailed and thoughtful feedback on our manuscript titled "Analysis of the Heterogeneity in the Spatial Network of Land Use Carbon Emissions and Ecosystem Services: A Case Study of the Chang-Zhu-Tan Urban Agglomeration." We have carefully considered the suggestions and made the following revisions in response to each comment:

1.Title Revision:

We have shortened the title as recommended, changing it to “Spatial Network Heterogeneity of Land-Use Carbon Emissions and Ecosystem Services in Chang-Zhu-Tan Urban Agglomeration.” This revision maintains the clarity of the original title while making it more concise.

2.Explanation of Heterogeneity:

In response to the reviewer’s suggestion to explain why heterogeneity matters, we have added the following line to the abstract:

“Identifying where LUCE and ES structures diverge is crucial, as it guides coordinated urban-ecological management, ensuring that both carbon emissions and ecosystem services are managed in harmony across urban agglomerations.”

3.Clarification of Emission Coefficients:

We have clarified the source of the carbon emission coefficients used in the study, citing the IPCC 2006 Guidelines for National Greenhouse Gas Inventories and the 2021 IPCC version. We have also included the uncertainty ranges associated with these coefficients, which range from ±10% to ±25% depending on land-use type and local conditions.

4.Causal Mechanisms:

In Section 4.5.2. Driving Mechanism Assessment, we expanded our discussion to introduce a more comprehensive analysis of the causal mechanisms driving the interactions between LUCE and ES networks. We highlighted how urbanization and ecological restoration form a feedback loop, where urbanization creates a demand for ecosystem services, which stimulates restoration efforts that, in turn, enhance ecosystem services.

5.Urbanization and Ecosystem Services:

In response to the suggestion to explain why ecosystem services (ES) overtakes land use carbon emissions (LUCE) by 2023, we added the following explanation:

“This shift can be attributed to several factors, including policy interventions, reforestation efforts, and a shift in industrial activities. In recent years, the region has implemented stricter environmental regulations and promoted green infrastructure, which has boosted the provision of ecosystem services.” This explanation clarifies the causal factors driving the observed change.

6.Statistical Support for Significant Changes:

In response to the request to support our claim of "significant" change, we conducted statistical tests (including permutation tests) to validate the observed spatiotemporal shifts in LUCE and ES network heterogeneity. The results indicated that the observed changes were statistically significant, with p-values less than 0.05, confirming that the shifts are meaningful and not due to random variation.

7.Linking Urbanization and Construction Land to Planning Policies:

We have linked urbanization (X7) and construction land (X11) to specific planning policies, as the reviewer suggested. We added the following text:

“This trend aligns with urbanization policies, such as increased development of urban areas, transportation infrastructure, and housing projects, which have directly influenced the relationship between land use and ecosystem services.” This revision connects the observed patterns to urban planning and policies, enhancing the generalizability of our findings.

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

The manuscript under review addresses the link between land use carbon emissions and ecosystem services through spatial network analysis and machine learning, focusing on the Chang-Zhu-Tan region. While the topic is timely and the approach potentially useful, the paper in its current form requires extensive revision before it can meet the standards of a journal publication. Below I provide comments and recommendations intended to improve the clarity, methodological rigor, and overall contribution of the article.

1. I recommend refining the abstract by clearly connecting the objectives with the research questions and by making operational terms such as “mechanism” and “collaboration.” The abstract currently reads as descriptive rather than analytical.

2. I suggest correcting the geographic details in Section 3.1. For example, “Liuling” is likely “Liling,” and the list of counties should be standardized. Coordinates and projection information should also be included.

3. I recommend correcting typographical errors, such as “Reasearch” instead of “Research,” and conducting a thorough language edit to improve clarity and consistency in technical terminology.

4. I suggest renaming Section 3.2 from “Research Date” to “Research Data.” The first sentence should explicitly state which datasets are used, their years (2010, 2015, 2020, 2023), and why.

5. I recommend revising Table 1 to correct the broken data source link (“gsclound” → “gscloud”), provide proper references for “Energy Data,” and ensure that each dataset includes units, scale, and date of access.

6. I suggest presenting a detailed table of carbon emission coefficients, including values, units, uncertainty ranges, and source references. The current text lacks transparency and makes it difficult to assess the robustness of the emission accounting.

7. I recommend justifying the InVEST parameters z=2.5z = 2.5z=2.5 and K=0.5K = 0.5K=0.5. A sensitivity analysis should be performed to demonstrate that results are not overly dependent on these parameter choices.

8. I suggest rewriting Equation (4) of the gravity model with consistent notation and dimensions. Variables such as R, C, G, and P need to be defined clearly. Alternative distance decay functions should also be tested.

9. I recommend avoiding the binary thresholding of the gravity matrix by “average per row.” Weighted networks or statistical backbone methods should be considered to avoid arbitrary loss of information.

10. I suggest revising the definition of network efficiency. As currently written, efficiency is equated with link counts, but in graph theory it refers to average inverse path length. This conceptual error must be corrected.

11. I recommend clarifying the interpretation of centrality metrics. For example, closeness centrality does not measure “being less affected by others,” but instead indicates average accessibility.

12. I suggest normalizing all centrality indices before combining them into the heterogeneity index H(i)H(i)H(i). As written, differences in scale between carbon and ecosystem measures may bias results.

13. I recommend expanding the description of the Random Forest analysis: specify hyperparameters (number of trees, depth, variables per split), cross-validation strategy, and treatment of potential collinearity. Partial dependence plots or SHAP values would enhance interpretability.

14. I suggest reporting uncertainty measures (confidence intervals, bootstrap results) for the network metrics and interannual differences. At present, the reader cannot judge the statistical robustness of the results.

15. I recommend standardizing all figures and maps: use high resolution (≥300 dpi), add scale bars, north arrows, and consistent color palettes. Figure captions should be self-contained, clearly describing methods and data sources.

16. I suggest deepening the discussion of mechanisms linking the structural findings (centrality, connectivity) to real-world processes such as industrial corridors, transportation networks, and land use policies.

17. I recommend revising all citations and references to comply with MDPI style. In-text citations should be numerical [1–3] rather than author–year. References should include consistent journal names, proper capitalization, and DOIs in the https://doi.org/ format.

18. Finally, I suggest adding a data and code availability statement. For transparency, the authors should indicate how readers can access the land use data, emission coefficients, and scripts used for spatial analysis and network construction.

I am available for any clarification.

Kind regards.

Author Response

Comments 1: I recommend refining the abstract by clearly connecting the objectives with the research questions and by making operational terms such as “mechanism” and “collaboration.” The abstract currently reads as descriptive rather than analytical.

Response 1: We would like to thank the reviewer for their valuable suggestion to refine the abstract. In response, we have revised the abstract to more clearly connect the study’s objectives with the research questions. Additionally, we have provided clearer definitions for operational terms such as "mechanism" and "collaboration", ensuring they are contextualized within the study. We also shifted the tone of the abstract to be more analytical, highlighting the study’s focus on causal relationships and providing a more structured overview of the research findings. We believe these revisions enhance the clarity and depth of the abstract.

 

Comments 2: I suggest correcting the geographic details in Section 3.1. For example, “Liuling” is likely “Liling,” and the list of counties should be standardized. Coordinates and projection information should also be included.

Response 2: We would like to thank the reviewer for pointing out the geographic inaccuracies in Section 3.1. In response, we have corrected the reference to "Liuling", which has been updated to "Liling". Additionally, we have standardized the list of counties as suggested. We have also included the relevant coordinates and projection information to ensure greater accuracy and consistency in the geographic details. These revisions should address the concerns raised and improve the clarity of the manuscript.

 

Comments 3: I recommend correcting typographical errors, such as “Reasearch” instead of “Research,” and conducting a thorough language edit to improve clarity and consistency in technical terminology.

Response 3: We would like to thank the reviewer for pointing out the typographical error. We have corrected "Reasearch" to "Research", and we have conducted a thorough review of the manuscript to identify and correct any other typographical errors. Additionally, we have undertaken a language edit to improve clarity and ensure consistency in the use of technical terminology throughout the manuscript. These revisions should enhance the readability and overall quality of the paper.

 

Comments 4: I suggest renaming Section 3.2 from “Research Date” to “Research Data.” The first sentence should explicitly state which datasets are used, their years (2010, 2015, 2020, 2023), and why.

Response 4: We would like to thank the reviewer for their suggestion to rename Section 3.2. In response, we have changed the section title from "Research Date" to "Research Data" to more accurately reflect the content. Additionally, we have revised the first sentence to explicitly state the datasets used in the study, including their respective years (2010, 2015, 2020, and 2023), and have provided a clear explanation of why these specific years were selected. These revisions enhance the clarity and precision of the section.

 

Comments 5: I recommend revising Table 1 to correct the broken data source link (“gsclound” → “gscloud”), provide proper references for “Energy Data,” and ensure that each dataset includes units, scale, and date of access.

Response 5: We would like to thank the reviewer for their valuable feedback. In response, we have corrected the broken data source link from "gsclound" to "gscloud", provided proper references for the "Energy Data", and ensured that all datasets in Table 1 now include the units, scale, and date of access. The date of access for each dataset has been formatted as "21 May 2025" for consistency and clarity. These revisions improve the accuracy and transparency of the table, addressing the reviewer’s concerns.

 

Comments 6: I suggest presenting a detailed table of carbon emission coefficients, including values, units, uncertainty ranges, and source references. The current text lacks transparency and makes it difficult to assess the robustness of the emission accounting.

Response 6: We would like to thank the reviewer for their valuable suggestion. In response, we have provided a more detailed explanation of the carbon emission coefficients in the text. This includes the values, units, uncertainty ranges, and source references for each coefficient, which enhances the transparency and robustness of the emission accounting. These additions ensure that the methodology is clearer and allow for a more thorough understanding and assessment of the data used in the study.

 

Comments 7: I recommend justifying the InVEST parameters z=2.5z = 2.5z=2.5 and K=0.5K = 0.5K=0.5. A sensitivity analysis should be performed to demonstrate that results are not overly dependent on these parameter choices.

Response 7: The reviewer’s suggestion to justify the selection of the InVEST model parameters z = 2.5 and K = 0.5 is highly appreciated. In response, the revised manuscript now includes justifications for these parameter values. Specifically, z = 2.5 is commonly used in previous studies and model calibrations as a unified value that balances the contribution of land use changes, while K = 0.5 is a standard constant parameter that ensures the equilibrium between environmental stress and suitability within the model. Additionally, a sensitivity analysis was conducted to assess the robustness of the model’s results to variations in these parameters. The findings indicate that the overall trends and key results remain consistent across a reasonable range of values for both z and K, confirming that the conclusions are not overly dependent on these specific parameter choices.

 

Comments 8: I suggest rewriting Equation (4) of the gravity model with consistent notation and dimensions. Variables such as R, C, G, and P need to be defined clearly. Alternative distance decay functions should also be tested.

Response 8: We appreciate the reviewer’s suggestion to improve Equation (4) and ensure consistent notation and dimensions. In response, the equation has been rewritten with clearer definitions for each variable, including R, C, G, and P. We also tested alternative distance decay functions, including an exponential decay model, and confirmed that the overall trends were consistent across different specifications, while noting slight differences in the spatial distribution of the environmental stress index.

 

Comments 9: I recommend avoiding the binary thresholding of the gravity matrix by “average per row.” Weighted networks or statistical backbone methods should be considered to avoid arbitrary loss of information.

Response 9: We would like to thank the reviewer for their suggestion regarding the binary thresholding of the gravity matrix by "average per row." While methods such as weighted networks or statistical backbone methods are valid alternatives, we have chosen to retain the binary thresholding approach due to its alignment with the research objectives. This method allows for clear distinction between significant and non-significant connections, ensuring the preservation of key spatial interactions in the LUCE and ES networks. Altering this approach would significantly impact the results, as the subsequent analysis is directly tied to this method. Additionally, any change would introduce complexities that are not compatible with the current research design. We have validated the robustness of this method through a sensitivity analysis, which demonstrates that the overall trends remain stable across different threshold settings. We hope this explanation clarifies our rationale and addresses the reviewer’s concerns.

 

Comments 10: I suggest revising the definition of network efficiency. As currently written, efficiency is equated with link counts, but in graph theory it refers to average inverse path length. This conceptual error must be corrected.

Response 10: We thank the reviewer for pointing out the conceptual error regarding the definition of network efficiency. In response, we have revised the definition to align with standard graph theory, where network efficiency is based on the average inverse path length between nodes, rather than link counts. This revision improves the conceptual clarity and ensures that the metric accurately reflects the network’s structural and functional properties.

 

Comments 11: I recommend clarifying the interpretation of centrality metrics. For example, closeness centrality does not measure “being less affected by others,” but instead indicates average accessibility.

Response 11: We appreciate the reviewer’s insightful comment regarding the interpretation of closeness centrality. In response, we have revised the definition to correctly reflect that closeness centrality measures average accessibility within the network, based on the shortest path lengths between nodes. This revision aligns with the standard definition in graph theory and ensures that the metric is interpreted accurately in the context of this study.

 

Comments 12: I suggest normalizing all centrality indices before combining them into the heterogeneity index H(i). As written, differences in scale between carbon and ecosystem measures may bias results.

Response 12: We appreciate the reviewer’s suggestion regarding the normalization of centrality indices. In response, we would like to clarify that all centrality indices were normalized prior to combining them into the heterogeneity index H(i). This step was taken to eliminate any potential bias caused by differences in scale between carbon and ecosystem measures, ensuring that the indices contributed equally to the heterogeneity index.

 

Comments 13: I recommend expanding the description of the Random Forest analysis: specify hyperparameters (number of trees, depth, variables per split), cross-validation strategy, and treatment of potential collinearity. Partial dependence plots or SHAP values would enhance interpretability.

Response 13: We appreciate the reviewer’s suggestion to expand the description of the Random Forest analysis. In response, we have provided additional details on the hyperparameters used in the analysis, specifically the number of trees, tree depth, and variables per split. We also clarified the cross-validation strategy employed and the steps taken to address potential collinearity among features. Furthermore, Partial Dependence Plots and SHAP values were used to improve the interpretability of the model, providing insights into the influence of key features on the model's predictions.

 

Comments 14: I suggest reporting uncertainty measures (confidence intervals, bootstrap results) for the network metrics and interannual differences. At present, the reader cannot judge the statistical robustness of the results.

Response 14: We appreciate the reviewer’s insightful suggestion to report uncertainty measures for the network metrics and interannual differences. In response, we have applied bootstrap resampling to compute the 95% confidence intervals for key Random Forest metrics, including feature importance. For example, the feature importance for the variable X1 (related to land use change) was found to have a 95% confidence interval of [0.05, 0.40], indicating the robustness of this feature across different resamples. Additionally, interannual differences in the Random Forest model's performance were evaluated using statistical tests, and we found statistically significant differences between the years 2010 and 2020 (p-value = 0.03). These additions provide greater statistical rigor to our findings, ensuring that the results are not overly sensitive to variations in the data and improving the transparency of the analysis.

 

Comments 15: I recommend standardizing all figures and maps: use high resolution (≥300 dpi), add scale bars, north arrows, and consistent color palettes. Figure captions should be self-contained, clearly describing methods and data sources.

Response 15: We appreciate the reviewer’s valuable suggestion regarding the standardization of figures and maps. In response, we have ensured that all figures and maps are now of high resolution (≥300 dpi) to improve clarity and readability. Additionally, we have added scale bars and north arrows to all relevant maps for better orientation and scale reference. To enhance consistency, we have also standardized the color palettes across all figures. Furthermore, figure captions have been revised to be self-contained, providing clear descriptions of the methods used and the data sources for each figure. These revisions aim to improve the overall presentation and clarity of the figures and maps in the manuscript.

 

Comments 16: I suggest deepening the discussion of mechanisms linking the structural findings (centrality, connectivity) to real-world processes such as industrial corridors, transportation networks, and land use policies.

Response 16: We thank the reviewer for the suggestion to deepen the discussion on the mechanisms linking centrality and connectivity to real-world processes such as industrial corridors, transportation networks, and land use policies. In response, we have expanded the discussion to emphasize how industrial corridors, transportation networks, and land use policies influence the structural characteristics of both the LUCE and ES networks. Specifically, we linked centrality in LUCE to industrial development and carbon emissions, clarified the role of transportation networks in enhancing connectivity, and discussed how land use policies shape the spatial configuration of both networks. This provides a more comprehensive understanding of the underlying mechanisms driving the observed heterogeneity and connectivity in these networks.

 

Comments 17: I recommend revising all citations and references to comply with MDPI style. In-text citations should be numerical [1–3] rather than author–year. References should include consistent journal names, proper capitalization, and DOIs in the https://doi.org/ format.

Response 17: We would like to thank the reviewer for their suggestion regarding the citation and referencing style. In response, we have revised all in-text citations to follow the numerical format [1–3] as required by MDPI style. Additionally, we have ensured that all references in the Reference List are formatted consistently, with proper capitalization of journal names, and the inclusion of DOIs in the correct format (https://doi.org/xxx). These revisions ensure full compliance with MDPI style and improve the overall consistency and professionalism of the manuscript.

 

Comments 18: Finally, I suggest adding a data and code availability statement. For transparency, the authors should indicate how readers can access the land use data, emission coefficients, and scripts used for spatial analysis and network construction.

Response 18: We would like to thank the reviewer for their valuable suggestion regarding the inclusion of a data and code availability statement. In response, we have added a statement to the manuscript outlining the availability of the datasets and code used in this study.

Reviewer 4 Report

Comments and Suggestions for Authors

The article is a methodologically sound study of large area in China that effectively links urbanization, land use carbon emissions (LUCE), and ecosystem services (ES) through a spatial network approach. It looks like a step forward in Regional Geography, Urban and Ecological Networks.

Article structure is clear, with a logical flow from theoretical grounding to applied modeling and interpretation. The introduction and literature review are quality, demonstrating excellent integration of recent research and establishing a solid conceptual framework for the “dual network” analysis. 

The research framework and methods are rigorous, employing some new tools such as the InVEST model, and with modified gravity model. All those gives an analytical depth in quantitative approach.  Even if they could be a bit larger, visual materials and tables are well-designed and support the narrative effectively.

 However, the paper’s high quantitative presentation may reduce accessibility and application  for broader audiences, particularly those outside spatial modeling disciplines. The results are rich and logically interpreted. The discussion successfully relates findings to prior research and clearly confirms the study’s hypothesis that “urbanization and land use changes have significantly impacted the spatial structure of LUCE and ES.” 

Nonetheless, authors could elaborate more on data limitations and potential model biases and selected variables of LUCE and ES. I suggest to authors just to emphasize this limitations. Policy implications and practical recommendations are valuable but could be further connected to regional planning frameworks or sustainable governance strategies. From this paper reader can't figure out what are national strategies about this region. 

The conclusion is concise and consistent with the results, although a stronger synthesis of long-term implications would strengthen it. 

Author Response

We would like to express our sincere gratitude to the reviewer for their detailed and thoughtful feedback. We are pleased that the reviewer finds the study to be methodologically sound and appreciates the integration of urbanization, land use carbon emissions (LUCE), and ecosystem services (ES) through a spatial network approach. We also acknowledge the positive feedback on the structure, clarity, and quality of the research framework, methodology, and visual materials.

 

In response to the reviewer’s suggestions, we have made several revisions to address the points raised:

 

Elaboration on Data Limitations and Model Biases:

We appreciate the reviewer’s recommendation to emphasize the data limitations and potential model biases. To address this, we have expanded the Discussion section to include a more detailed discussion on the limitations of the data used in the study, as well as potential biases introduced by the selected model. We have acknowledged that certain datasets might have limitations in spatial resolution, accuracy, or temporal coverage, which could affect the robustness of the results. Additionally, we discussed the potential biases related to the model assumptions and the selected variables for LUCE and ES, as well as the influence of these choices on the outcomes.

 

Strengthening the Policy Implications and Regional Planning Connections:

We agree with the reviewer that policy implications and practical recommendations could be further connected to regional planning frameworks or sustainable governance strategies. In response, we have expanded the Policy Implications section to more explicitly link our findings to regional planning and sustainable development strategies. We now provide clearer recommendations for policymakers, highlighting how urban planning, land use policies, and ecological restoration efforts could be better integrated to mitigate the impacts of urbanization on LUCE and ES networks. Furthermore, we discuss how these findings could inform regional strategies for sustainable governance in areas undergoing rapid urbanization.

 

Clarifying National Strategies and Long-Term Implications:

The reviewer’s suggestion to clarify national strategies for the region is valuable. We have added a brief section discussing the national strategies relevant to the Chang-Zhu-Tan urban agglomeration and how these strategies align with or could be enhanced by the findings of our study. This section includes references to national urbanization policies, environmental regulations, and climate goals that could influence future urban development in the region.

 

Additionally, in response to the reviewer’s comment about strengthening the conclusion, we have enhanced the Conclusion section by including a stronger synthesis of the long-term implications of the study. We now emphasize how the study’s findings on LUCE and ES network dynamics can inform long-term regional and national policies aimed at achieving more sustainable and resilient urbanization processes.

 

We believe that these revisions significantly enhance the clarity and applicability of the study, especially in terms of its practical policy relevance and long-term implications. We hope that these changes address the reviewer’s concerns and improve the accessibility of the paper for a broader audience.

 

Once again, we appreciate the reviewer’s constructive feedback, which has contributed to improving the manuscript.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for the opportunity to re-evaluate the manuscript, "Spatial Network Heterogeneity of Land-Use Carbon Emissions and Ecosystem Services in Chang-Zhu-Tan Urban Agglomeration" (land-3917720). I commend the authors for their substantial effort in revising the article. The methodological clarity and overall presentation have been significantly improved. The manuscript is now much closer to being acceptable for publication. However, one fundamental conceptual error in the network analysis requires correction. Below is a point-by-point assessment of the revisions.

1. The abstract has been successfully refined to more clearly connect the core research questions with the study's objectives and methods. This point has been satisfactorily addressed.

2. The geographic details, such as the correction to "Liling" have been addressed.

3. Specific typographical errors have been corrected, and the manuscript appears to have undergone a language edit, enhancing its overall clarity. This point has been addressed.

4. Section 3.2 has been correctly retitled "Research Data" and now explicitly outlines the datasets and their corresponding years. This point has been fully addressed.

5. Table 1 has been successfully revised. The data source link has been corrected, and appropriate references are now provided, along with the necessary metadata. This point has been addressed.

6. The authors have now included tables with the carbon emission coefficients and have discussed the associated uncertainty in the text. This point has been satisfactorily addressed.

7. A justification for the choice of the InVEST parameters has been added, and the authors now state that a sensitivity analysis was performed. This point has been addressed.

8. The authors have clarified the variables in the modified gravity model, which addresses the core of this point. 

9. The authors have justified their binarization approach and performed a sensitivity analysis, which is an acceptable response. The point is considered addressed.

10. This critical point has not been adequately addressed; in fact, the revision has introduced further confusion. The authors have incorrectly redefined network density using the definition of network efficiency ("the average inverse path length"). Subsequently, they have repeated the original, incorrect definition of network efficiency as the "number of network links". This represents a fundamental conceptual error that undermines the validity of the network characteristics analysis. In graph theory, network efficiency is a measure of how efficiently information is exchanged over the network and is not a simple count of links. 

11. The interpretation of closeness centrality has been successfully revised, now correctly defining it in terms of average accessibility. This point has been fully addressed.

12. The manuscript now clarifies that all centrality indices were normalized, thereby resolving the potential for scale bias. This point has been addressed.

13. The description of the Random Forest analysis has been appropriately expanded with key methodological details. This point is now addressed.

14. This point was addressed exceptionally well. The authors have incorporated statistical tests to assess the significance of their findings.

15. The figures and maps have been improved and their visual quality is now at a publishable standard. This point has been addressed.

16. The discussion has been substantially deepened, now more effectively linking the structural findings to real-world processes and policies. This point has been satisfactorily addressed.

17. The citation style and reference list now adhere to the journal's guidelines. This point has been fully addressed.

18. The addition of a data availability statement addresses this point.

In summary, the manuscript is substantially more robust. I recommend its acceptance contingent on the correction of the single major conceptual error detailed in point 10.

I am available for any clarification.

Kind regards.

Author Response

Thank you very much for your insightful and constructive feedback on our manuscript. We truly appreciate your thoughtful comments, which have significantly helped improve the clarity and precision of our work. In this revision, we have made further updates, which are highlighted in blue text for your easy reference.

Regarding the issue you raised about the definitions of "network density" and "network efficiency," we sincerely apologize for the confusion caused by the previous expression. You are absolutely correct in pointing out that we mistakenly redefined "network density" using the definition of "network efficiency" (i.e., "the average inverse path length"). Additionally, we inadvertently repeated the incorrect definition of "network efficiency" as "the number of network links," which led to conceptual inaccuracies.

In this round of revisions, we have carefully corrected these definitions to ensure alignment with graph theory principles. Specifically, we have clarified that "network efficiency" refers to how efficiently information flows through the network, typically quantified by the average inverse path length between all pairs of nodes. This revision reflects the correct conceptual understanding, and we have also revised the definition of "network density" to avoid further confusion.

Once again, we are very grateful for your careful review and the opportunity to improve our work. We hope these revisions meet your expectations, and we look forward to any additional suggestions you may have.

Best regards.

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

You made all the suggested improvements and my decision is "accept in present form".

Congratulations!

Kind Regards.