Rapid Emergency Response Resilience Assessment of Highway Bridge Networks under Moderate Earthquakes

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper presents a methodology to evaluate the post-disaster emergency response capacities and provide a prompt seismic assessment of regional bridges. The study introduces several original contributions:

- Development of a novel emergency response resilience vector for highway bridge networks.

- Application of a feedforward neural network model as a substitute for Incremental Dynamic Analysis (IDA) in building fragility curves.

- Implementation of a decision tree-based approach for simulating emergency responses after earthquakes.

- Integration of transportation engineering, structural engineering, and performance-based earthquake engineering (PBEE).

Compared to other published materials, this paper introduces an interdisciplinary approach that combines finite element methods, intelligent algorithms, and traffic flow simulations to provide a comprehensive and rapid assessment of the seismic resilience of highway bridge networks. This approach is not commonly found in existing studies and offers a practical and scalable solution for emergency response planning and resilience evaluation.

The conclusions are consistent with the evidence and arguments presented in the paper. The study's findings are supported by the case study of the Sioux Falls HBN, where the proposed methodologies were applied and validated. The paper effectively addresses the main questions through the development and application of the ANN-based seismic assessment method and the decision tree-based emergency response model.

Additionally, the authors should address the following points:

1. In Figure 1, the "H" in "HBN simulation" is missing.

2. Provide more details about the geometry of the studied bridges used to train the ANN.

3. Further explain the parameters of Equation (4), particularly the definitions of (t_0), (t), and (t_h) based on the survival function.

4. Clarify if, in the IDA analysis of the ANN model trained in Section 3.2, the earthquakes were scaled and if the set of seismic motion signals were tested in each bridge configuration or if the configurations were taken at random.

5. Discuss the computational cost of the IDA results compared to the ANN model of Section 3.2.

6. Specify where the earthquake loads were applied in the finite element model. Were displacements or accelerations imposed?

7. Confirm if it is possible to obtain fragility curves from the results obtained with the ANN and whether the differences between the fragility curves obtained with IDA and with ANN can be visualized.

8. It is commendable that the authors have made the database used in the historical emergency incident dataset calculations available on GitHub.

9. Provide a clearer explanation of what Figure 6b represents.

10. Offer more details on the durability model used for the results at 10, 20, 30, and 40 years.

Finally, the authors might consider implementing sensitivity analyses to understand the impact of the input parameters on the resilience assessment outcomes and identify the most relevant features for the ANN model.

 

Comments on the Quality of English Language

The English language is appropriate. 

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Overall, the quality of English is good, however there is room for improvement in some points.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript is detailed and well organized, however lacks some important information. The few comments are listed below

 Introduction:

Lines 48-52: Lack of specificity about what "highly intricate" means in terms of seismic responses.

Line 53-57: Needs clearer connection between traditional methods and the limitations they present for region-scale analysis. Including a comparison with modern methods can provide more context.

Line 64-66: The benefits of using artificial neural networks and decision trees are mentioned but need more detailed justification or examples.

Intelligent Resilience Analysis Framework of HBN

Lines 108-110: Lacks a clear description of Fig. 1, which should be briefly explained.

Lines 121-124: Lacks details on how the functionality vector is constructed and applied.

Line 160: Consider defining spectral acceleration (Sa) more clearly.

Rapid Seismic Assessment of Regional Bridges

The simplified methods suggested to replace FEM-based techniques might not fully capture the complexities and nuances of bridge responses to seismic events. This could lead to inaccurate assessments of bridge safety and functionality post-earthquake. Please explain.

Justification for certain methodological choices (e.g., K=10 for cross-validation) is lacking.

The threshold of 0.01% for discontinuing large-scale emergency rescue might be arbitrary and needs to be based on empirical data.

Intelligent Emergency-Response Decision

Lines 391-394: The inconsistent predictions are attributed to rescue efforts and random factors, but this analysis is superficial. There is no discussion on how to improve the model or handle these inconsistencies.

Lines 411-415: Clearly define assumptions for better understanding.

Lines 421-430: The description of the graph model for HBNs is somewhat vague. More detail is needed on how nodes and edges are defined, and how traffic capacities and times are measured and updated.

Application to HBN Case Studies

Line 482: The analysis assumes the Sioux Falls network is located in a second-tier city, but it lacks justification or context for this choice. Providing reasoning for this assumption would enhance the credibility of the analysis.

Lines 517-527: While the method for evaluating seismic motion intensity is mentioned, the explanation lacks detail. More information on the Probability Seismic Hazard Analysis (PSHA) method and the Ground Motion Prediction Equation (GMPE) model would be beneficial.

Lines 553-575: Additional discussion on the implications of the analysis results and potential recommendations would enhance the section.

Conclusion

A section on the limitations of the study is missing. Acknowledging these can provide a balanced view and identify areas for future research. Rewrite the conclusion section in detail.

Comments on the Quality of English Language

simplify and clarify the language in dense sections. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Excellent work!!!

Author Response

The authors of the manuscript thank the works of the reviewers firstly. With the comments, the authors have the possibility to improve the quality of the manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have not attended any of my comments. The must have misplaced the comments of other reviewers to mine. I would like to request them to kindly check and resend again.

Author Response

Please see the attachment.

The authors are very sorry for the mistakes. 

Author Response File: Author Response.docx

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

All my comments have been attended well. No further queries.