Simulation of Metro Congestion Propagation Based on Route Choice Behaviors Under Emergency-Caused Delays

Round 1

Reviewer 1 Report

The article submitted researches congestion propagation during peak-period commuting considering passengers’ routing preferences. The article is well written and can be understood however here are my concerns:

I suggest more detail in explaining for foreigners the scenario in which the research has been conducted (maybe with figure or tables). Which is the frequency of metros, how long do stops (commuter hop on and off) usually take, if there are screen/panels announcing how long will it take to fix the emergency/delay, etc. It is important to know if commuters leave after having really waited for 10 minutes or if there is a screen warning of a 10 minute delay.

Regarding section 3.1, A further explanation of the data collected is needed. Table1 does not give any information about what knowledge is extracted from the surveys and further paragraphs only mention age, travelling purpose and that a 52% of travellers can only endure waiting for less than 10 minutes. The document's experiment leans on this data so a more in deep explanation is required for a better understanding if the article's conclusions are relevant or not.

In regard to the methodology of sections 3.2 and 3.3 it does not describe how these theoretical considerations connect with the data and variables gathered in the surveys. The theoretical considerations presented in these sections are then introduced in the route choice models but again a more extensive explanation of the collected data would help connecting which variables go into formulas and then into the choice model.

The methodology and simulation models are well explained and constructed. Figure 11 requires more granularity to see how the scenario evolves between t=0s and t=20s. However would the entire model be suitable for real-time simulation for incident
management? How long does it take to extract meaningful results from the model since it is stochastic and must be run several times?

Finally, I suggest extracting some conclusions and running some different emergency scenarios. You have only described your methodology but no mention to the knowledge extracted. I believe you can add some quantitative results about how long does it take to recover from emergencies or the effects of different time consuming emergencies you may have tested. In addition, it would be desirable to add some comparison of the obtained results with other similar research works.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

1.The delay time of route 2-4 in Table 1 why is 0 minutes? please explain the survey assumption.

2.The variables coefficient of ASC 1, ASC 2, ASC 3, ASC4, ASC 5, and Transfer times in Table 2 why are positive. please explain the reason.

3.What is AC/m in Table 2? Please describe and explain the meaning.

4.Why there are no the SIR-Exposed curve in Figure 6? Please describe and explain.

5.Line 216 of " many variables could influence passenger's mood and routes references" or "many variables could influence passenger's mode and routes references", please check again which one is better.

Author Response

Please see the attachment.

Author Response File: Author Response.docx