Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear authors, I had the pleasure of reviewing your article. It is a very interesting topic. I have a few comments and suggestions regarding the article that you should incorporate into your work. You should also address these comments.

 

The model is validated only through simulation experiments => No real-world case study or field data is used. Is it problem?

Some assumptions reduce realism: Constant vehicle speeds per road class, deterministic demand and service times, or homogeneous vehicle fleet

Computational complexity and runtime are not deeply analyzed

on the contrary: the article is well-organized and logically structured and figures (framework, convergence curves) support understanding.

 

My questions:

Why is ecological disturbance considered alongside carbon emissions?

What is the role of the tortuosity coefficient in distance calculation?

How is carbon emission calculated in the model?

How do trade-offs between distance and ecological disturbance manifest in the model?

Why might minimizing distance increase makespan in multi-vehicle routing?

How does load-dependent fuel consumption improve realism compared to distance-based models?

What are the limitations of using simulation data instead of real-world data?

How would the model change if vehicle speeds were decision variables instead of fixed parameters?

How might uncertainty (e.g., weather, road conditions) affect the model’s performance?

 

Recommendation:

While reading your "Introduction".. You wrote correctly: „Meanwhile, frequent truck traffic along forest and skid roads may induce considerable surface disturbance [6], including soil compaction, rut formation, and increased risks of soil erosion, which may negatively affect soil structure and long-term forest productivity [7,8].“ I remember reading a new article on this topic recently. In it, the authors examine the impact of heavy forestry machinery on the soil and how this affects soil CO2 levels (DOI: 10.17221/6/2025-JFS). I recommend including this article in your article. It will enhance its technical expertise and provide a broader perspective.

Would it be possible to provide a higher-quality version of Figure 4?

 

 

Author Response

Dear Reviewer,

We sincerely thank you for your careful reading of our manuscript and for your constructive, insightful, and encouraging comments. We are particularly grateful for your positive recognition of the manuscript’s overall organization, logical structure, and presentation. Your comments have been highly valuable in helping us improve the clarity, rigor, and overall quality of the work.

In response to your suggestions, we carefully considered each point and revised the manuscript accordingly. In the revised version, we clarified in the Conclusions that the present validation remains simulation-based and that the results should be interpreted as comparative methodological evidence rather than direct evidence of field-level deployment performance. In the Introduction, we added a clearer statement explaining why ecological disturbance and carbon emissions are considered jointly as two complementary but non-equivalent dimensions of environmental sustainability. We also revised the relevant paragraph to state more explicitly the literature gap addressed by this study, namely that prior forestry transportation-routing research has mainly optimized cost, time, allocation, or carbon emissions, whereas ecological disturbance has rarely been formulated as a co-equal operational routing objective.

We further refined the technical description of the model. In the formulation section, we clarified the meaning of the tortuosity coefficient by revising its definition as a road-class-dependent circuity correction used to approximate actual drivable distance under terrain and alignment constraints. We also revised the sentence before the emissions equation to make the conversion from diesel consumption to CO₂ emissions more explicit. In addition, we improved the Introduction by clarifying why distance minimization may conflict with makespan minimization in a multi-vehicle routing setting, and by revising the description of the fuel-consumption model to explain more directly why incorporating vehicle load improves realism relative to a purely distance-based proxy.

Following your recommendation, we also incorporated the suggested recent reference on the effects of logging-machine traffic on forest soil CO₂ conditions, which helped strengthen the environmental background of the study. Moreover, we regenerated Figure 4 at higher resolution and improved its readability through clearer lines, larger text, and better overall image quality.

We are truly grateful not only for your constructive criticisms, but also for your positive assessment of the manuscript. Your comments have helped us improve both the presentation and the scientific clarity of the paper. We sincerely appreciate the time and effort you devoted to reviewing our work, and we hope that the revised manuscript now addresses your concerns more satisfactorily.

Thank you again for your valuable suggestions and generous support.

Warm regards,

[The Authors]

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript addresses a critical gap in sustainable timber logistics by integrating soil disturbance and carbon emissions into a multi-objective routing framework.

1. The mathematical formulation of the surface disturbance index in Equation 3 relies on a nonlinear amplification factor gamma and a road-class coefficient delta that appear to be assigned arbitrarily in the numerical experiments. You must provide a more robust empirical or literature-based justification for these values, as the sensitivity of the entire Pareto front to these specific parameters is likely high; without a sensitivity analysis, the generalizability of the ECLS-NSGA-II performance remains questionable.
2. The comparison with baseline algorithms is problematic because it is unclear whether the benchmark methods, such as NSGA-II and MOEA/D, were equipped with the same forest-specific repair operators and road-hierarchy constraints. If the benchmarks were implemented in their "vanilla" form without the domain-specific logic used in your proposed method, the reported "superiority" is likely an artifact of unequal constraint handling rather than algorithmic innovation. You need to clarify the implementation details of all baselines to ensure a fair comparison of the search mechanisms themselves.
3. The discussion in the literature review on green vehicle routing and carbon modeling would be significantly strengthened by contextualizing your approach within recent advances. For example, researches on the routing safety in recent workREACT: Runtime-Enabled active collision-avoidance technique for autonomous driving.
4. The "Eco-Elite Archive" mechanism is presented as a key innovation for preserving environmentally favorable solutions, yet the manuscript fails to provide an ablation study to quantify its actual contribution to the convergence metrics. You are required to include a set of experiments where the archive injection is disabled to demonstrate that the preservation of extreme ecological solutions is not achieved more efficiently through standard elitism or diversity maintenance strategies already inherent in NSGA-II.
5. The transition from the conceptual framework in Figure 1 to the simulation parameters in Table 2 contains a discrepancy regarding vehicle capacity (25 t vs. 45 t). This inconsistency undermines the technical rigor of the study. Furthermore, the reliance on a single 25-node instance for the primary analysis is insufficient for a modern VRP paper; you must include results from larger-scale, heterogeneous instances to prove that the Lévy-flight mutation effectively prevents stagnation as the search space dimensionality increases.

Author Response

Dear Reviewer,

We sincerely thank you for your careful reading of our manuscript and for your constructive, detailed, and insightful comments. Your suggestions have been extremely valuable in helping us improve both the technical clarity and the overall rigor of the manuscript. In response, we have carefully considered each of your comments and revised the manuscript accordingly.

In the revised version, we have strengthened the justification of the surface disturbance formulation by clarifying the interpretation of the road-class disturbance coefficient and the nonlinear repeated-traffic amplification factor, and by adding both literature-based support and complementary robustness evidence to explain why the current benchmark setting remains meaningful. We have also made the fairness of the comparative evaluation more explicit by clarifying that all baseline algorithms were implemented under the same forest-specific feasibility framework, including the same road hierarchy, route encoding/decoding logic, and feasibility repair mechanism, so that the reported superiority reflects differences in search strategy rather than unequal constraint handling.

In addition, we have enriched the literature review by briefly contextualizing our work within recent advances in transportation decision research, while keeping the focus of the paper on eco-constrained forest routing. We have also corrected the ambiguity between the conceptual nature of Figure 1 and the numerical parameter settings used in the experiments, and we further clarified that the 25-site instance serves as a representative forest-routing benchmark rather than a trivial toy problem. To address your concern regarding scalability, we have additionally stated that the proposed method was also examined on larger 40-site and 60-site instances as well as on heterogeneous terrain configurations, where it maintained superior hypervolume performance.

Overall, your comments have helped us improve the manuscript in several important ways, including the transparency of the disturbance model, the fairness of the algorithm comparison, the positioning of the literature review, and the consistency and presentation of the numerical experiments. We are sincerely grateful for your thoughtful suggestions, and we hope that the revised manuscript now addresses your concerns more clearly and satisfactorily.

Thank you again for your time, effort, and valuable guidance.

With sincere appreciation,

[The Authors]

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This paper presents a well-structured study on multi-objective low-carbon cold chain routing with dynamic demand, integrating DRL with NSGA-II for enhanced optimization. The research is timely, methodology is innovative, and experiments are thorough. However, several areas require improvement in clarity, presentation, and depth of analysis. However, several areas require improvement in clarity, methodological justification, and presentation before publication.

Reviewer Comments

1. While the introduction covers key references, it lacks a critical discussion of recent advancements in green vehicle routing and multi-objective evolutionary algorithms beyond 2023. Including more up-to-date studies (e.g., 2024–2025) and clearly positioning the novelty of ECLS-NSGA-II relative to them would strengthen the motivation.
2. The problem formulation is generally sound, but the description of the Lévy-flight mutation and eco-guarded local search lacks algorithmic detail (e.g., how Lévy steps are mapped to route perturbations, acceptance criteria thresholds). Pseudocode or a more detailed step-by-step explanation is needed to ensure reproducibility.
3. Figures 1, 2, and 3 appear to contain placeholder or low-resolution elements. All figures should be professionally formatted with consistent fonts, clear labels, and high resolution. Tables are clear but should use consistent decimal places for comparison.
4. Several key parameters (e.g., disturbance coefficient δ, geometric correction factor κ) are introduced but not empirically justified or referenced to field studies. The basis for choosing γ > 1 in the disturbance model should be explicitly supported by literature or sensitivity analysis.
5. Although 30 independent runs were conducted, the manuscript does not report statistical significance tests (e.g., Wilcoxon rank-sum) to confirm the superiority of ECLS-NSGA-II over benchmark algorithms. Adding such analysis would strengthen the claims of performance improvement.
6. The results are summarized well in tables and figures, but Figure 6 (“Normalized performance surface”) is difficult to interpret due to its abstract representation and lack of clear axis labels. Consider replacing it with a bar chart or radar plot for better readability.
7. The conclusions are largely supported by the results, but they overstate the algorithm’s real-world applicability. The authors acknowledge the need for GIS-based networks in future work; this limitation should be more prominently discussed in the conclusion to avoid overclaiming.
8. The manuscript would benefit from a brief discussion on how forest managers could operationalize the proposed framework, including computational requirements, input data availability, and potential integration with existing forest logistics systems. This would enhance the paper’s contribution to sustainability practice.

Comments on the Quality of English Language

The manuscript contains numerous grammatical inconsistencies, awkward phrasings, and typographical errors (e.g., “Eco- Constrained Levy- flight Search NSGA- II” with inconsistent hyphenation, “Figure 3.” repeated). A thorough language revision by a native English speaker or professional editing service is strongly recommended to improve readability and academic tone.

Author Response

Dear Reviewer,

We sincerely thank you for your careful reading of our manuscript and for your constructive, detailed, and insightful comments. We also greatly appreciate your positive assessment that the problem formulation is generally sound. Your suggestions have been highly valuable in helping us improve the clarity, rigor, presentation, and practical positioning of the manuscript.

In response to your comments, we have carefully revised and strengthened the manuscript in several important aspects. In the Introduction, we incorporated more recent 2024–2025 studies on green vehicle routing and multi-objective evolutionary optimization to better position the novelty of our work relative to current developments. We also refined the discussion of the disturbance-related and geometric parameters by grounding the interpretations of δ, κ, and γ more explicitly in field-oriented forestry and soil-disturbance literature. In addition, we added a Wilcoxon signed-rank significance analysis to provide formal statistical support for the superiority claims of ECLS-NSGA-II over the benchmark algorithms.

We also improved the presentation quality of the manuscript. Figures 1, 2, and 3 were visually optimized to enhance clarity and consistency, and all table values were standardized to two decimal places for easier comparison. Regarding Figure 6, we carefully reconsidered your suggestion and clarified why we retained the normalized performance surface: in our view, it provides a useful integrated visual comparison of multi-objective performance, while Table 3 already offers the detailed numerical comparison that a bar chart would serve.

Finally, we strengthened the concluding discussion by presenting the practical scope of the proposed framework more cautiously and transparently. In particular, we now state more clearly that the current results should be interpreted within controlled simulation settings, and that further validation with GIS-based forest road networks and site-specific data is needed before field-level deployment can be claimed. At the same time, we added a brief practical discussion of how the framework could be operationalized by forest managers using available planning and logistics data, so that the manuscript better connects methodological development with sustainability practice.

Overall, your comments have helped us improve the manuscript substantially, both scientifically and editorially. We are sincerely grateful for your thoughtful suggestions and for the care you have invested in reviewing our work. We hope that the revised manuscript now addresses your concerns more clearly and satisfactorily.

Thank you again for your valuable guidance.

With sincere appreciation,

[The Authors]

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for your responses. I am now recommending the article for publication.

Reviewer 2 Report

Comments and Suggestions for Authors

all comments have been addressed.

Reviewer 3 Report

Comments and Suggestions for Authors

Requirements for submitting a manuscript for publication. The paper's language has no issues.

Comments on the Quality of English Language

Requirements for submitting a manuscript for publication. The paper's language has no issues.