How Determinants Affect Transfer Ridership between Metro and Bus Systems: A Multivariate Generalized Poisson Regression Analysis Method

Comment 0:

This study focuses on the transfer ridership between bus and metro systems under different dates and severe weather conditions, to quantify the impacts of various attributes on the transfer ridership of different transfer modes (metro-to-bus and bus-to-metro). A multivariate generalized Poisson regression (GPR) model is applied to investigate the effects of critical factors on the transfer ridership of different transfer modes on weekdays, holidays, and typhoon weather. However, some minor changes in the research should be modified. Therefore, I recommend reconsideration of the manuscript following minor revision.

Response:

Thank you for your helpful comments. We are very appreciative of your comments. All the suggested comments have been thoroughly examined and accordingly taken into account in our revision.

We hope these revisions meet your expectations.

Comment 1:

In line 174, the author mentions the gaps of relationship research between weather and transfer ridership. This paragraph should be written more concisely and structured of the factors impacting transfer ridership. It is a bit hard to read and get the key points.

Response:

Thanks for your comments. The literature review of this paper consists of two main parts: the transfer-related studies in Section 2.1 and the effects of weather on transport ridership in Section 2.2.

In the existing relevant studies, there are a few studies on the influence of transfer ridership. Therefore, this study refers to much literature on the analysis of factors influencing transport ridership. Among many studies, we have found that weather variables have significant impacts on transport ridership. So this paper also analyzes the impact of each weather variable on transfer ridership. Moreover, the studies related to the impact of each factor on transfer are analyzed and summarized in Section 2.1, and Table 1.

In line 174, the main function of this paragraph mainly summarizes the limitations of existing studies based on the previous literature descriptions.

We have revised the paragraph, and rewritten the sentence in line 174. Please check Lines 177-179 on Page 4 of the resubmitted manuscript.

For the reviewer’s convenience, the specific revisions in the resubmitted manuscript are given below:

 “Therefore, inspired by existing studies on the effect of weather on transport ridership, this study also analyzes the effect of weather variables on transfer ridership. However, there are still some limitations, mainly in two respects.”

We hope the explanations can answer your question.

Comment 2:

In the paragraph from line 219, the author demonstrated passengers’ shift from bus travels to metro travels. If so, how does this factor impact the research results? The author should give some further explanation to this.

Response:

Thank you for your comments. We have carefully revised the manuscript and added discussions and explanations of the effect of each factor on the research results.

 In the paragraph from line 219, this section mainly introduces the geographical features, weather characteristics, and the distribution of the public transportation network in Shenzhen. The reasons for choosing Shenzhen as the study area are further explained.

This study demonstrated passengers’ shift from bus travel to metro travel. We analyzed the impact of each factor on the transfer ridership of the metro-to-bus mode and the bus-to-metro mode. How these relevant factors affect the transfer ridership is also discussed in the results section. Moreover, we also give some further explanations for this.

(1) On weekdays and weekends, transfer ridership of the bus-to-metro mode is significantly affected by multiple factors, including rainfall, revised transfer time, house rent, housing prices, geographic GDP, crowd density, feeder bus routes, CBD distance, morning peak, evening peak, weekends. Among them, house rent, housing prices, and feeder bus routes are positively associated with the transfer ridership of the bus-to-metro mode. While other variables are negatively associated with the transfer ridership of the bus-to-metro mode. Please check lines 577-605 on Pages 20-21 of the resubmitted manuscript, to further explain the above phenomenon more clearly, we added the following contents in the resubmitted manuscript:

“On weekdays and weekends, for the bus-to-metro mode, crowd density has the most significant impact on transfer ridership, about 1.26 to 10.64 times that of other variables. Among the weather variables, only rainfall will affect transfer ridership. And heavy rainfall will increase the transfer ridership. The revised transfer time has a positive impact on transfer ridership. This is because the revised transfer time includes the in-vehicle time of taking the first bus, so the longer the revised transfer time is, the more transfer ridership is. Feeder bus routes are negatively correlated with transfer ridership. Because the next trip for transfer passengers is the metro. Therefore, the metro and the bus routes compete, and the fewer bus routes are, the more transfer passengers will choose the metro.

 Socioeconomic and population variables have a significant influence on transfer ridership. Among them, the house rent and housing prices are negatively associated with the transfer ridership. While the geographic GDP and crowd density are positively associated with the transfer ridership. It may be because people in stations with higher house rent and housing prices usually have higher incomes and travel mostly by taxi or private car, rarely relying on public transportation. Metro stations with higher geographic GDP are usually economically developed and have better public transportation networks, making it more convenient for people to travel and transfer, with higher transfer ridership. The CBD distance has a positive impact on the ridership. This is ascribed that metro stations are far from the CBD, the less developed economic level, and people who travel medium-long distances usually transfer mainly by public transportation, so there is more transfer ridership. In addition, the date variable has a significantly positive impact on transfer ridership. This is consistent with the distribution of transfer ridership on weekdays in Figure 5. Because the travel of commuters on weekdays is mainly concentrated in the morning and evening peak, and the morning and evening peak travel path direction is just the opposite. Therefore, more people transfer from bus to metro in the morning peak, while they transfer from metro to bus in the evening peak.”

(2) On holidays, transfer ridership of the bus-to-metro mode is significantly affected by all independent factors, including weather variables, transfer-related variables, socioeconomic and population variables, and Built environment variables. Among them, visibility, house rent, housing prices, and feeder bus routes are negatively associated with the transfer ridership of the bus-to-metro mode. While other variables are positively associated with the transfer ridership of the bus-to-metro mode. Please check lines 621-652 on Pages 21-22 of the resubmitted manuscript, to further explain the above phenomenon more clearly, we added the following contents in the  resubmitted manuscript:

“On holidays, for the bus-to-metro mode, among all variables, housing prices have the most significant effect on transfer ridership, approximately 1.19 to 5.75 times that of other variables. The housing prices are negatively associated with transfer ridership. Weather variables have a significant impact on transfer ridership. Among them, the maximum temperature, rainfall, and maximum wind speed have a positive impact on the transfer ridership, whereas the minimum visibility has the opposite effect. This shows that under the conditions of high temperature, heavy rainfall, strong wind, or low visibility, more travelers choose to transfer from the metro to the bus, leading to an increase in the transfer ridership. Moreover, among the weather variables, rainfall has the greatest impact on transfer ridership, approximately 22.12% to 57.74% higher than other weather variables. The revised transfer time has a significant positive impact on transfer ridership. Feeder bus routes are negatively associated with transfer ridership. These findings are similar to the factors influencing transfer ridership on weekdays.

Similar to the study cases on a weekday, socioeconomic and population variables have a significant impact on transfer ridership. Among them, house rent and housing prices have a positive impact on the transfer ridership, whereas the geographic GDP and crowd density have the opposite effect. CBD distance is positively associated with the transfer ridership. The farther the metro station is from the CBD, the more transfer passengers are.”

(3) During the typhoon, transfer ridership of the bus-to-metro mode is significantly affected by multiple factors, including temperature, wind speed, visibility, rainfall, house rent, housing prices, geographic GDP, crowd density, feeder bus routes, and CBD distance. Among them, visibility, geographic GDP, crowd density, and CBD distance are positively associated with the transfer ridership of the bus-to-metro mode. While other variables are negatively associated with the transfer ridership of the bus-to-metro mode. Please check lines 679-697 on Pages 23-24 of the resubmitted manuscript, to further explain the above phenomenon more clearly, we added the following contents in the resubmitted manuscript:

“During the typhoon, for the bus-to-transfer mode, among all variables, the crowd density has the greatest influence on transfer ridership, about 1.27 to 7.74 times that of other variables. Which is positively associated with the transfer time. This indicates that crowd density can facilitate transfer ridership at metro stations. Unlike the case of the metro-to-bus mode, temperature, wind, and rainfall have a remarkably negative effect on the transfer ridership, whereas visibility has the opposite impact. Moreover, among the weather variables, wind speed has the greatest impact on transfer ridership, about 1.26 to 3.89 times that of other weather variables. This indicates that the transfer ridership is sensitive to changes in weather conditions. This may be ascribed to the fact that the first trip of the bus-to-metro mode is taking a bus. During typhoon days, in extremely bad weather (strong wind speed and rainstorm conditions), few passengers choose to take the bus for the first journey, so there is less transfer ridership. Similar to the cases of weekdays and holidays, the revised transfer time has a considerably positive impact on the transfer ridership. Feeder bus routes are negatively associated with the transfer ridership.

Similar to the cases of weekdays and holidays, socioeconomic variables have a significant impact on transfer ridership. Among them, the house rent and housing prices have a negative impact on the transfer ridership, whereas the geographic GDP has the opposite effect. CBD distance is positively associated with the transfer ridership.”

We hope the explanations can answer your question.

Comment 3:

 Figure 2 is too complicate to visualize data, variables and models, and the author might need to use more figures instead one to show the hierarchy among these. Figure 3 and Fig 4 are hard to check meteorological conditions at the same, and the visualization method may need to be improved.

Response:

Thank you very much for your suggestion. We have carefully revised the manuscript, the corrections are as follows:

(1)   Figure 2 is too complicated to visualize data, variables, and models. Therefore, we redrew Figure 2 and changed the original Figure 2 to Figure 2 and Figure 3 in the resubmitted manuscript. Figure 2 and Figure 3 in the resubmitted manuscript represent the schematic diagram of the passenger transfer process and the hierarchy of the variables in the models, respectively. Please check lines 261-265 on Page 8 of the resubmitted manuscript, we added the following figures in the resubmitted manuscript:

Figure 2. The transfer process between the metro and the bus systems. (a) the transfer from metro to bus, (b) the transfer from bus to metro.

Figure 3. The hierarchy of the variables in the models.

(2)   Figure 3 and Figure 4 in the original manuscript are hard to check meteorological conditions at the same, but the visualization method has been improved. Therefore, we redrew Figure 3 and Figure 4 of the original manuscript. We plot the heat maps for the weather and transfer-related variables, as shown in Figure 4 and Figure 5 of the resubmitted manuscript. Please check lines 281-287 on Page 9 of the resubmitted manuscript, we added the following figures in the resubmitted manuscript:

Figure 4. The distribution of weather variables: (a) wind speed, (b) rainfall, (c) minimum visibility, and (d) maximum temperature.

“From Figure 4,  it is easy to check meteorological conditions at the same. It can be seen that the four weather variables, wind speed, rainfall, temperature, and visibility, have great fluctuations on the 14^th, 15^th, and 16^th of October, this is an extreme weather condition. Which is also typhoon weather. Therefore, it is necessary and meaningful to analyze the transfer ridership in typhoon weather separately.”

Please check lines 316-328 on Pages 10-11 of the resubmitted manuscript, we added the following figures in the resubmitted manuscript:

“To show more clearly the distribution of different transfer-related variables on different dates, we added heat maps of transfer ridership and transfer time, as shown in Figure 5 below:

Figure 5. The distribution of transfer ridership and transfer time: (a) the transfer ridership of the metro-to-bus mode, (b) the transfer ridership of the bus-to-metro mode, (c) the transfer time of the metro-to-bus mode, and (d) the revised transfer time of the bus-to-metro mode.

From Figure 5, it can be seen that the transfer ridership has great fluctuations on the 14^th, 15^th, and 16^th of October, with significant morning and evening peaks. Therefore, it is necessary and meaningful to analyze the transfer ridership in typhoon weather separately.”

 We hope these revisions can meet your expectations.

Comment 0:

The authors in this paper focused on the analysis of trips between bus and metro systems at different times and in severe weather conditions to determine the impact of attributes on trips on specific systems. The authors used a multivariate generalised Poisson regression (GPR) model to determine the impact.

The evident strengths of the work include:

- clear structure of the work;

- thorough literature review;

- original presentation of research results.

Response:

Thank you for your comments. We have carefully revised and proofread the manuscript. We hope these revisions can meet your expectations.

Comment 1:

Possible shortcomings of the work include:

- the scope of the study covers October 2017. Is this not too late given that we are currently in July 2022?

Response:

Thank you for your comments. The scope of the study covers October 2017. We mainly explored the effect of various variables such as weather variables on transfer ridership. Until July 2022, Shenzhen's residents rely primarily on public transportation for daily travel, and transfer ridership still accounts for a significant proportion of public transport ridership. The methodology used, and the important findings and conclusions obtained in this study, are still applicable to transfers in today's public transportation system. Certainly, the results obtained with the new data are more time-sensitive. Therefore, new data will be collected to further explore the effects of various factors on transfer ridership in the subsequent study.

We hope the explanations can answer your question.

Comment 2:

- in the study, it would be worthwhile to present studies confirming the results achieved or possibly to carry out studies periodically, e.g. in 4 seasons (this provision could be included in future plans)

Response:

Thank you for your suggestions. It would be worthwhile to present studies confirming the results achieved or possibly to carry out studies periodically. This provision is included in plans.

 Please check Lines 886-890 on Page 28 of the resubmitted manuscript. For the reviewer’s convenience, the specific revisions in the resubmitted manuscript are given below:

“It would be worthwhile to present studies confirming the results achieved or possibly to carry out studies periodically, e.g. in 4 seasons. Therefore, we will subsequently conduct a periodical study to explore the influence of various factors on transfer ridership during all seasons of the year, to derive the variability of transfer ridership affected by various factors in different seasons.”

 We hope the explanations can answer your question.

Comment 3:

- Shenzhen city has a moderately warm climate. On average, the least rainy days are in December (4.53 days). The month with the most rainy days is June (23.73 days). Consequently, isn't it worthwhile to carry out research in just these two extreme months?

Response:

Thank you for your comments. The month with the rainiest days is June (23.73 days). Consequently, it is worthwhile to carry out research in just these two extreme months. But, We currently do not have available data for these two months. Therefore, we studied the data for October, as this month has a great variability of weather, and also has extreme weather such as typhoons and rainstorms, which can meet our research needs.

Please check Lines 216-218 on Page 7 of the resubmitted manuscript. For the reviewer’s convenience, the specific revisions in the resubmitted manuscript are given below:

“As shown in Figure 1, the climatic conditions change greatly, and the weather in October is hot.”

Please check Lines 230-236 on Page 7 of the resubmitted manuscript. For the reviewer’s convenience, the specific revisions in the resubmitted manuscript are given below:

“The entire month of October 2017 was chosen as the study period in this study. There are two reasons for this: first, October 2017, with 16 workdays, eight holidays (including the National Day and Mid-Autumn Festival), and six ordinary weekends, meets the scope and requirements of this study. We can also capture the transfer ridership affected by various occasions besides the general travel characteristics. Second, the weather in October changes dramatically, there are strong typhoons and stormy weather is suitable for our research topics on exploring the impact of extreme weather on the transfer ridership.”

We hope the explanations can answer your question.

Comment 4:

Nevertheless, in my opinion the article is suitable for publication in its current form.

Response:

Thank you for your comments. We have revised the manuscript accordingly. We hope these revisions can meet your expectations.

Comment 0:

This study uses the smart card data to examine the relationship between the association between weather factors and the transfer ridership. Dependent variable in this study is the hourly transfer ridership. A count model, i.e., the generalized Poisson was applied.

The paper is generally well-written and structured. The results are clearly explained, while the findings are discussed in a well-balanced manner. I only have a few minor suggestions. These are outlined below in relation to the relevant sections of the paper.

Response:

Thank you for your comments. We have carefully revised and proofread the manuscript. We hope these revisions can meet your expectations.

Comment 1:

Section 3. Study Area and Data:

I am worried about the sample size. The dataset only contains one month information in 2017. This data set is relatively low and may introduce some random effects. Do the authors feel confident about the validity and generalizability of the findings? As mentioned before, Shenzhen is a fast-growing city, so I wonder, can one-month data from five years ago now accurately provide useful information?

Can the socio-demographic data obtained from smart card system? Are the data matched with the card holders? I am asking this because I think maybe the traveler's age, gender, occupation, etc., are also the important factors affecting the transfer?

For the weather information, raw weather data shall be obtain from different Meteorological stations in Shenzhen? Then, when waiting or transferring at each bus or metro station, are the weather data matched with the nearest Meteorological stations? For example, when I am at bus station A, the weather should be matched with the nearest weather station a. This is a common approach used in the studies on the effects of weather.

Response:

Thanks for your questions. Our revisions and answers are illustrated as follows:

1. Although we studied only one month of sample data, the sample still covered the metro and bus systems in Shenzhen. So the sample size of this paper is sufficient to support the research content of this study. We are confident in the validity and generalizability of the findings. Moreover, some existing related studies only used one month of sample data and the conclusions obtained are feasible [47, 54]. Shenzhen is a rapidly developing city, metro network and bus network of the city had been very well built five years ago, The new metro stations and bus stops built in the following years are mainly distributed in the suburbs. The essential factors affecting the transfer ridership are still the weather variables and the economic level of the city, the built environment, and other factors. Therefore, the use of data from one month five years ago can provide accurate and useful information. Furthermore, there are also relevant studies that use data from five years ago to analyze the impact of various factors on transfer [55]. The issue of study data is also explained in the resubmitted manuscript. Please check Lines 239-250 on Page 7 of the resubmitted manuscript:

“Although we studied only one month of sample data from October 2017, the sample still covered the metro and bus systems in Shenzhen. The sample size of this paper is sufficient to support the research content of this study. We are confident in the validity and generalizability of the findings. Moreover, some existing related studies also only used one month of sample data and the conclusions obtained are feasible [47,54]. Shenzhen is a rapidly developing city, five years ago metro network and bus network of the city has been very well built. The new metro stations and bus stops built in the following years are mainly distributed in the suburbs. The essential factors affecting the transfer ridership are still the weather variables and the economic level of the city, the built environment, and other factors. Therefore, the use of data from one month five years ago can provide accurate and useful information. Furthermore, there are also relevant studies that use data from five years ago to analyze the impact of various factors on transfer [55].”

2. We can count the population from the smart card data, and this data is the count of card holders. However, it is difficult to count the age, gender, and occupation of travelers, which are the profile data of travel behavior for individual travelers. These data involve the privacy of travelers. It is difficult to obtain these data. In a follow-up study, we will try to obtain the traveler's age, gender, and occupation, to further investigate the impact of these factors on transfer ridership.

3. For the weather information, raw weather data should obtain from different Meteorological stations in Shenzhen. Our weather data is also derived from aggregated data from various Meteorological stations in Shenzhen. The study area of this paper is mainly the downtown area of Shenzhen. The weather conditions in each metro station are similar. Therefore, this paper uses the same weather data at each metro station for the study. Moreover, the weather variables of the metro stations in different metro stations were studied using the same data in the relevant studies. For example, Miao et al. [41] and Zhou et al. [47] studied the effect of weather on public transport ridership, the weather data were the same at different metro stations.

We hope that the above illustrations can solve your questions.

References:

[ 41]   Miao, Q.; Welch, E.W.; Sriraj, P.S. Extreme Weather, Public Transport Ridership and Moderating Effect of Bus Stop Shelters. J. Transp. Geogr. 2019, 74, 125–133, doi:10.1016/j.jtrangeo.2018.11.007.

[ 47]   Zhou, M.; Wang, D.; Li, Q.; Yue, Y.; Tu, W.; Cao, R. Impacts of Weather on Public Transport Ridership: Results from Mining Data from Different Sources. Transp. Res. Part C Emerg. Technol. 2017, 75, 17–29, doi:10.1016/j.trc.2016.12.001.

[54]    Ma, X.; Zhang, J.; Ding, C.; Wang, Y. A Geographically and Temporally Weighted Regression Model to Explore the Spatiotemporal Influence of Built Environment on Transit Ridership. Comput. Environ. Urban Syst. 2018, 70, 113–124, doi:10.1016/j.compenvurbsys.2018.03.001.

[ 55]      Li, W.; Chen, S.; Dong, J.; Wu, J. Exploring the Spatial Variations of Transfer Distances between Dockless Bike-Sharing Systems and Metros. J. Transp. Geogr. 2021, 92, 103032, doi:10.1016/j.jtrangeo.2021.103032.

Comment 2:

3.2.3. Built environment variables, The authors may want to show the area of CBD of Shenzhen in Figure 1?

For Tables 3 4 5, VIF values can be removed by simply describe them in text. More importantly, since this study used a multivariate approach, please provide the correlation parameters between two dependent variables. Also, it is necessary to provide the model fit statistics to justify the final specification of the model.

For discussion section, I think there is lack of in-depth discussion on why these results are found. Especially when the focus is on the weather effects. For example, why the strong wind condition will contribute to the increase in the metro-to-bus transfer during the typhoon? It is more dangerous to take the bus (outdoor) while the indoor metro station is more safe, isn't it? or, why the high temperature will increase the metro-to-bus transfer? Similar to the first question, bus station is in outdoor while the waiting environment in Metro station is more comfortable based on previous studies. I just want to emphasize that more discussion should be made based on previous literature.

One important thing is to provide policy implication or suggestions based on the effects of weather on transfer ridership. Please elaborate more.

Response:

Thanks for your comments. We have carefully revised and proofread the manuscript. Our revisions and answers are illustrated as follows:

1. Figure 1 shows the geographical location, climate characteristics, and travel demand in Shenzhen, China. Figure 1 mainly shows the distribution of metro lines and metro stations, as well as the area and location of each administrative region. The core CBD of Shenzhen is in Futian District. The built environment variables mainly include the feeder distance of bus stops from the metro station and the distance of the metro station from the core CBD of Futian District.

2. For Tables 3 4 5, the variance inflation factor (VIF) is used to examine multicollinearity among independent variables. And many related kinds of literature, this indicator is shown in the result tables [23,42]. Moreover, the fourth reviewer thinks that the VIF values are important and need to be described in detail. Therefore, in tables 3 4 5, we kept this parameter. More importantly, since this study used a multivariate approach, we provide the correlation parameters between two dependent variables. We calculated and visualized the Pearson correlation coefficients between two variables in each model. Please check Lines 539-555 on Pages 18-19 of the resubmitted manuscript, the details are as follows:

“Different variables have different effects on the transfer ridership. Moreover, this study used a multivariate approach. To further illustrate the correlation between the variables in each model, we calculated Pearson correlation coefficients between the variables and visualized the coefficients as shown in Figures 10, Figure 11, and Figure 12 below.

Figure 10. The Pearson correlation coefficient of various variables on workdays and weekends. (a) the metro-to-bus mode, and (b) the bus-to-metro mode.

Figure 11. The Pearson correlation coefficient of various variables on holidays. (a) the metro-to-bus mode, and (b) the bus-to-metro mode.

Figure 12. The Pearson correlation coefficient of various variables on holidays. (a) the metro-to-bus mode, and (b) the bus-to-metro mode.

From Figure 10, Figure 11, and Figure 12, the correlation coefficients of the variables in each model are less than 1, and most of the correlation coefficients are less than 0.5. Therefore, the variables are independent of each other. Which satisfies the requirements of each model for the independent variables.”

Also, it is necessary to provide the model fit statistics to justify the final specification of the model. Therefore, in table 3, table 4, and table 5, we added diagnostic statistics of models. Please check Line 560 on Page 19, Line 653 on Page 22, and Line 674 on Page 23 of the resubmitted manuscript.

3. For the discussion section, The results section is discussed in depth again why these results are found. The reasons for the impact of these factors on transfer ridership are explained. Especially when the focus is on the weather effects. We explained the reasons why transfer ridership was affected by the weather factors. Research results show that strong winds, heavy rain, and high temperatures will increase transfer ridership of the metro-to-bus mode but reduce transfer ridership of the bus-to-metro mode. This result is counterintuitive. Because it is more dangerous to take the bus (outdoor) while the indoor metro station is safer. This may be described to the fact that transfer passengers travel in typhoon weather differ from that under normal weather conditions. To explain this phenomenon, we will further investigate the impact of other factors such as the travel purpose of transfer passengers on transfer ridership in bad weather. Moreover, among the weather variables, the temperature has the greatest impact on transfer ridership of the metro-to-bus mode, while wind speed has the greatest impact on transfer ridership of the bus-to-metro mode. This may be ascribed to the fact that transfer activities are usually exposed to the outdoors, and are more affected by the weather conditions. Moreover, the metro-to-bus mode has the opposite path to the bus-to-metro mode. This study focuses on the determinants of transfer ridership between metro and bus systems. Usually, transfer passengers cannot take a single vehicle directly to their destination and have to transfer. In bad weather, such as hot weather, people prefer to take public transportation over other transfer modes, such as shared bikes, electric bikes, or walking. Therefore, high temperatures will increase the transfer ridership. In other words, in bad weather, for those transfer passengers who took the metro for the first leg of their journey and they had to make a transfer to get to their destination, most of them will choose to take the bus to reach their destination on the second leg of their journey. As a result, the high temperature will increase the transfer ridership of the metro-to-bus mode.

The impacts of each factor on transfer ridership are discussed in depth and the reasons behind this are explained one by one.

(1) Please check Line Lines 712-740 on Pages 24-25 of the resubmitted manuscript. The details are as follows:

“From Figure 13, it can be seen that among the influencing factors, transfer time, socio-economics, and population have a significant impact on transfer ridership on weekdays and weekends. Peak hours positively influence the transfer ridership. Morning peak, evening peak, low housing prices, low house rents, and high crowd density can attract more transfer ridership at the metro stations, the opposite can attract less transfer ridership at the metro stations. This may be described that a large of commuters gathering during peak hours. The majority of commuters require transfers to reach their destination. The house rent and housing prices are negatively associated with the transfer ridership. It may be because people in stations with higher house rent and housing prices usually have higher incomes and travel mostly by taxi or private car, rarely relying on public transportation. Crowd density is positively associated with transfer ridership. Because metro stations with high pedestrian density usually have more transfer passengers.

Additionally, the impact of some factors on transfer ridership varies under different transfer modes. For example, temperature and CBD distance have a significantly positive effect on transfer ridership of the metro-to-bus mode. This is ascribed that metro stations are far from the CBD, the less developed economic level, and people who travel medium-long distances usually transfer mainly by public transportation, so there is more transfer ridership. In high temperatures, people prefer to transfer by public transportation than other transfer modes such as biking or walking. While the feeder bus routes negatively influence transfer ridership of the bus-to-metro mode. Because the next trip for transfer passengers is the metro. Therefore, the metro and the bus routes compete, and the fewer bus routes are, the more transfer passengers will choose the metro. Transfer time has a significant negative impact on transfer ridership of the metro-to-bus mode. While the revised transfer time has a significant positive impact on transfer ridership of the bus-to-metro mode. Moreover, the former has a greater impact on transfer ridership than the latter, about 2.68 times as much as the latter. This is because the revised transfer time includes the in-vehicle time of taking the first bus, so the longer the revised transfer time is, the more transfer ridership is. The transfer passengers from metro to bus wish to take the bus earlier with the shorter transfer time, which can attract more transfer passengers.”

(2) Please check Line Lines 744-776 on Pages 25-26 of the resubmitted manuscript. The details are as follows:

“From Figure 14, it is clear that weather, built environment, and socioeconomic and demographic variables have significant effects on transfer ridership on holidays. Among them, housing prices and crowd density have the most significant impact on transfer ridership. Compared to the transfer ridership on weekdays, transfer ridership on weekends is more susceptible to weather factors. This is because passengers on holiday are more flexible in terms of travel time and purpose, and many will change their travel plans because of weather changes, such as canceling trips, changing their travel destinations, or changing their travel mode. Low housing prices, low house rents, high crowd density, heavy rain, and long-distance from CBD can attract more transfer ridership at the metro stations. Besides, the impact of some factors on transfer ridership differs in different transfer modes. For instance, temperature positively affects the transfer ridership of the metro-to-bus mode. It may be because people in stations with higher house rent and housing prices usually have higher incomes and travel mostly by taxi or private car, rarely relying on public transportation. Crowd density is positively associated with transfer ridership. Because metro stations with high pedestrian density usually have more transfer passengers. This is ascribed that metro stations are far from the CBD, the less developed economic level, and people who travel medium-long distances usually transfer mainly by public transportation, so there is more transfer ridership. While visibility negatively influences transfer ridership of the metro-to-bus mode. This may be because the lower the visibility is, the fewer passengers choose to walk, or bike and the more passengers choose public transportation to transfer. Transfer time has a significant negative impact on transfer ridership of the metro-to-bus mode, while the revised transfer time has a significant positive impact on transfer ridership of the bus-to-metro mode. This is because the revised transfer time includes the in-vehicle time of taking the first bus, so the longer the revised transfer time is, the more transfer ridership is. The transfer passengers from metro to bus wish to take the bus earlier with the shorter transfer time, which can attract more transfer passengers. Feeder bus routes have a positive impact on the transfer of the metro-to-bus mode, while feeder bus routes have a negative impact on the transfer ridership of the bus-to-metro mode. Because, the next trip for transfer passengers of the bus-to-metro mode is the bus, there are more bus routes, and the more transfer passengers are. For the bus-to-metro mode, the next trip for transfer passengers is the metro. Therefore, the metro and the bus routes compete, and the fewer bus routes are, the more transfer passengers will choose the metro.”

(3) Please check Line Lines 780-811 on Pages 26-27 of the resubmitted manuscript. The details are as follows:

“From Figure 15, it can be seen that socioeconomic and demographic variables, weather variables, and the built environment have significant effects on transfer ridership during typhoon weather. Among them, housing prices have the most significant impact on transfer ridership. Low housing prices, low house rents, high crowd density, and long distance from CBD can attract more transfer ridership at the metro stations. It may be because people in stations with higher house rent and housing prices usually have higher incomes and travel mostly by taxi or private car, rarely relying on public transportation. Crowd density is positively associated with transfer ridership. Because metro stations with high pedestrian density usually have more transfer passengers. This is ascribed that metro stations are far from the CBD, the less developed economic level, and people who travel medium-long distances usually transfer mainly by public transportation, so there is more transfer ridership. Besides, the impact of some factors on transfer ridership varies with transfer modes. Notably, the three weather variables have completely different effects on the transfer ridership for the two transfer modes. Strong winds, heavy rain, and high temperatures will increase transfer ridership of the metro-to-bus but reduce transfer ridership of the bus-to-metro. This result is counterintuitive. Because it is more dangerous to take the bus (outdoor) while the indoor metro station is safer. This may be described to the fact that transfer passengers travel in typhoon weather differ from that under normal weather conditions. To explain this phenomenon, we will further investigate the impact of other factors such as the travel purpose of transfer passengers on transfer ridership. Moreover, among the weather variables, the temperature has the greatest impact on transfer ridership of the metro-to-bus mode, while wind speed has the greatest impact on transfer ridership of the bus-to-metro mode. This may be ascribed to the fact that transfer activities are usually exposed to the outdoors, and are more affected by the weather conditions. The metro-to-bus mode has the opposite path to the bus-to-metro mode. Furthermore, in bad weather, such as hot weather, people prefer to take public transportation over other transfer modes, such as shared bikes, electric bikes, or walking. Therefore, high temperatures will increase the transfer ridership. In other words, in bad weather, for those transfer passengers who took the metro for the first leg of their journey and they had to make a transfer to get to their destination, most of them will choose to take the bus to reach their destination on the second leg of their journey. As a result, the high temperature will increase the transfer ridership of the metro-to-bus mode.”

We hope that these revisions in the manuscript can meet your expectations, and these explanations can solve your question about the results.

References:

[23]   Yan, X.; Levine, J.; Zhao, X. Integrating Ridesourcing Services with Public Transit: An Evaluation of Traveler Responses Combining Revealed and Stated Preference Data. Transp. Res. Part C Emerg. Technol. 2019, 105, 683–696, doi:10.1016/j.trc.2018.07.029.

[42]    Wei, M.; Liu, Y.; Sigler, T.; Liu, X.; Corcoran, J. The Influence of Weather Conditions on Adult Transit Ridership in the Sub-Tropics. Transp. Res. Part A Policy Pract. 2019, 125, 106–118, doi:10.1016/j.tra.2019.05.003.

Comment 0:

The paper is clear, well written, and meets the standard for scientific contribution both in format and in content. However, there are several issues that authors can work on:

Response:

Thank you for your comments. We have carefully revised and proofread the manuscript.

We hope these revisions can meet your expectations.

Comment 1:

1.     Line 282 and 284, upper bound of 40 mins and 50mins, the author should explain how those thresholds were chosen.

Response:

Thanks for your comments. The upper bound of 40 minutes and 50 minutes are mainly set regarding the literature related to recognizing transfer methods [9-10]. Existing survey studies show that 95% of the transfer ridership of the metro-to-bus mode have transfer times within 40 minutes, while 95% of the transfer ridership of the bus-to-metro mode have transfer times within 50 minutes [59]. Therefore, for the initial identification of the transfer ridership, we set the upper bound of 40 minutes and 50 minutes as the elapsed time thresholds for the metro-to-bus mode and bus-to-metro mode, respectively. We also explain how those thresholds were chosen in the resubmitted manuscript. Please check Lines 296-307 on Page 10 of the resubmitted manuscript:

“The method for identifying transfer time and transfer ridership is derived from the literature [10,56,57] and has been improved based on the following aspects. First, we record the travel time of all passengers every hour from the metro station to a candidate bus station, which is less than the upper bound of 40 min. We also record the travel time of all passengers every hour from the bus stop to a candidate metro station that is less than the upper bound of 50 min. Specifically, the upper bound of 40 minutes and 50 minutes are mainly set regarding the literature related to recognizing transfer methods [10,58]. Existing survey studies show that 95% of the transfer ridership of the metro-to-bus mode have transfer times within 40 minutes, while 95% of the transfer ridership of the bus-to-metro mode have transfer times within 50 minutes [59]. Therefore, for the initial identification of the transfer ridership, we set the upper bound of 40 minutes and 50 minutes as the elapsed time thresholds for the metro-to-bus mode and bus-to-metro mode, respectively.”

We hope that the above illustrations can explain how those thresholds were chosen and solve your question.

References:

[9]     Seaborn, C.; Attanucci, J.; Wilson, N.H.M. Analyzing Multimodal Public Transport Journeys in London with Smart Card Fare Payment Data. Transp. Res. Rec. 2009, 55–62, doi:10.3141/2121-06.

[10]  Huang, Z.; Xu, L.; Lin, Y.; Wu, P.; Feng, B. Citywide Metro-to-Bus Transfer Behavior Identification Based on Combined Data from Smart Cards and GPS. Appl. Sci. 2019, 9, doi:10.3390/app9173597.

[56]   Zhao, D.; Wang, W.; Woodburn, A.; Ryerson, M.S. Isolating High-Priority Metro and Feeder Bus Transfers Using Smart Card Data. Transportation (Amst). 2017, 44, 1535–1554, doi:10.1007/s11116-016-9713-7.

 [57]  Zhao, D.; Wang, W.; Li, C.; Ji, Y.; Hu, X.; Wang, W. Recognizing Metro-Bus Transfers from Smart Card Data. Transp. Plan. Technol. 2019, 42, 70–83, doi:10.1080/03081060.2018.1541283.

[58]    Wu, P.; Ph, D.; Xu, L.; Ph, D.; Li, J.; Ph, D.; Guo, H.; Ph, D.; Huang, Z.; Ph, D. Recognizing Real-Time Transfer Patterns between Metro and Bus Systems Based on Spatial-Temporal Constraints. 2022, 148, doi:10.1061/JTEPBS.0000721.

[59]    Gordon, J.B.; Koutsopoulos, H.N.; Wilson, N.H.M.; Attanucci, J.P. Automated Inference of Linked Transit Journeys in London Using Fare-Transaction and Vehicle Location Data. Transp. Res. Rec. 2013, 2343, 17–24, doi:10.3141/2343-03.

Comment 2:

For figure 4,5,6, I suggest to modify the label of the x-axis to clearly indicate the numbers are referring to the time (hour) of a day. Also, consider adding (a1)(a2)(a3)….to the subplots.

Response:

Thanks for your comments. For Figure 6,7,8 of the resubmitted manuscript, we have modified the label of the x-axis to indicate the numbers are referring to the time (hour) of a day. Also, we have added (a1)(a2)(a3)….to the subplots. Please check Lines 338-342 on Page 11, Lines 359-364 on Page 12, and Lines 384-389 on Page 13 of the resubmitted manuscript. For the convenience of the reviewer, the specific revisions in the resubmitted manuscript are given below:

Figure 6. The distribution characteristics of metro ridership: (a1), (a2), (a3) denote metro ridership on weekdays, weekends, and holidays, respectively; (b1), (b2), (b3) denote inbound ridership on weekdays, weekends, and holidays, respectively; (c1), (c2), (c3) denote outbound ridership on weekdays, weekends, and holidays, respectively.

Figure 7. The distribution characteristics of the transfer ridership: (a1), (a2), (a3) denote transfer ridership of metro stations on weekdays, weekends, and holidays, respectively; (b1), (b2), (b3) denote transfer ridership of the metro-to-bus mode on weekdays, weekends, and holidays, respectively; (c1), (c2), (c3) denote transfer ridership of the bus-to-metro mode on weekdays, weekends, and holidays, respectively.

Figure 8. The distribution characteristics of bus ridership and transfer time: (a1), (a2), (a3) denote bus ridership on weekdays, weekends, and holidays, respectively; (b1), (b2), (b3) denote transfer time of the metro-to-bus mode on weekdays, weekends, and holidays, respectively; (c1), (c2), (c3) denote revised transfer time of the bus-to-metro mode on weekdays, weekends, and holidays, respectively.

We hope that the figures revised in the manuscript can meet your expectation.

Comment 3:

3.     The transfer ridership in this paper is measured in term of the number of transfer passengers from metro to bus per hour and the number of transfer passengers from bus to metro per hour, this is mentioned in Table 2, but it would be better to clarify this at the beginning of the paper to help readers understand what the author is trying to model at the first place.

Response:

Thanks for your comments. In the introduction section, we clarify the definition of transfer ridership at the beginning of the paper to help readers understand what the author is trying to model in the first place. Please check Lines 50-55 on Page 2 of the resubmitted manuscript:

“Accordingly, this study focuses on the transfer between metro and bus systems. The definitions of the two transfer modes, namely metro-to-bus and bus-to-metro, can be found in [9]. The transfer ridership of the metro-to-bus mode in this paper is measured in terms of the number of transfer passengers from metro to bus per hour. And the transfer ridership of the bus-to-metro mode in this paper is measured in terms of the number of transfer passengers from bus to metro per hour.”

We hope these revisions can meet your expectations.

References:

[9]     Seaborn, C.; Attanucci, J.; Wilson, N.H.M. Analyzing Multimodal Public Transport Journeys in London with Smart Card Fare Payment Data. Transp. Res. Rec. 2009, 55–62, doi:10.3141/2121-06.

Comment 4:

4.     In line 419, “Therefore, to eliminate the influence of different dimensions on the results, all continuous independent variables are normalized”. The author should explain how the data normalization is conducted, is it scaling all variables to the same scale, if so, how it is scaled and what is the range after scaling? Also, are all the variables normally distributed? If some variables have highly skewed distribution, log transformation might be considered to normalize the distribution.

Response:

Thanks for your comments. To eliminate the influence of different dimensions on the results, all continuous independent variables are normalized. The calculation formula is as follows:

Where  denotes the normalized value,  denotes the raw value,  is the maximum value of the variable sample,  is the minimum value of the variable sample. It is scaling all variables to the same scale. The range of the variable after scaling is from 0 to 1. Moreover, after normalization, all the variables are normally distributed.

 Please check Lines 454-467 on Pages 15-16 of the resubmitted manuscript.

Hope our answer can solve your question. We hope that these revisions in the manuscript can meet your expectations,

Comment 5:

In Table 2, the unit of Minimum visibility per hour is m/s, this is not consistent with the definitions.

Response:

Thanks for your comments. In Table 2, the unit of Minimum visibility per hour is m. Please check Line 452 on Page 15 of the resubmitted manuscript.

Hope our answer can solve your question.

Comment 6:

In Table 3, 4, 5. does the “Observations” under Diagnostic statistics refers to the number of observations? Also, please clearly define all the terms like VIF. 

Response:

Thanks for your comments. In Tables 3, 4, and 5. the “Observations” under Diagnostic statistics refers to the number of observations. The variance inflation factor (VIF) is used to examine multicollinearity among independent variables. If VIF values are less than 5, it indicates that there is no multicollinearity between the independent variables of models [23,42]. In this study, as shown in Table 3, Table 4, and Table 5. All VIF values are less than 4, suggesting that no multicollinearity exists among the independent variables. Please check Lines 519-524 on Page 18 of the resubmitted manuscript. For the convenience of the reviewer, the specific revisions in the resubmitted manuscript are given below:

“The variance inflation factor (VIF) is used to examine multicollinearity among independent variables. If VIF values are less than 5, it indicates that there is no multicollinearity between the independent variables of models [23,42]. The corresponding results are presented in Table 4, Table 4, and Table 5. The level of the variance inflation factors (VIF) is calculated, and all are less than 4, suggesting that no multicollinearity exists among the independent variables.”

We hope these revisions can answer your question.

References:

[23]   Yan, X.; Levine, J.; Zhao, X. Integrating Ridesourcing Services with Public Transit: An Evaluation of Traveler Responses Combining Revealed and Stated Preference Data. Transp. Res. Part C Emerg. Technol. 2019, 105, 683–696, doi:10.1016/j.trc.2018.07.029.

[42]    Wei, M.; Liu, Y.; Sigler, T.; Liu, X.; Corcoran, J. The Influence of Weather Conditions on Adult Transit Ridership in the Sub-Tropics. Transp. Res. Part A Policy Pract. 2019, 125, 106–118, doi:10.1016/j.tra.2019.05.003.

Comment 7:

line 493 “For the metro-to-bus mode, it can be seen that the effect of transfer time on transfer ridership is the most significant among all independent variables, about 1.22 to 35.45 times that of other variables.” How those numbers are calculated should be clarified., Looks like it is derived based on the estimated coefficients: 2.127/0.06 = 35.45.

Also, normally directly comparing the coefficients is not very meaningful, because variables are in different unit and scale. Therefore, when comparing the coefficients, the author should emphasize that the variables have been normalized into the same scale or using the term “scandalized coefficient” instead of influence coefficient to avoid confusions.

The author could also consider using some other better metrics like marginal effects to compare the importance of different variables instead of using the estimated coefficients.

Response:

Thanks for your comments. To eliminate the influence of different dimensions on the results, all continuous independent variables are normalized. The calculation formula is as follows:

Where  denotes the normalized value,  denotes the raw value,  is the maximum value of the variable sample,  is the minimum value of the variable sample. It is scaling all variables to the same scale. The range of the variable after scaling is from 0 to 1. Moreover, after normalization, the impact of each factor in the regression models on transfer ridership can then be compared. Therefore, when comparing the coefficients, the variables have been normalized to the same scale.

Line 565-566, “For the metro-to-bus mode, it can be seen that the effect of transfer time on transfer ridership is the most significant among all independent variables, about 1.22 to 35.45 times that of other variables.” Those numbers are derived based on the estimated coefficients: 2.127/0.06 = 35.45. In this study, we compare the impact of each factor on the transfer ridership by the estimated influencing coefficients. In the subsequent study, we also consider using some other better metrics like marginal effects to compare the importance of different variables instead of using the estimated coefficients. For the convenience of the reviewer, the specific revisions in the resubmitted manuscript are given below:

(1) Please check Lines 457-467 on Pages 15-16 of the resubmitted manuscript:

“To eliminate the influence of different dimensions on the results, all continuous independent variables are normalized. To eliminate the influence of different dimensions on the results, all continuous independent variables are normalized. The calculation formula is as follows:

Where  denotes the normalized value,  denotes the raw value,  is the maximum value of the variable sample,  is the minimum value of the variable sample. It is scaling all variables to the same scale. The range of the variable after scaling is from 0 to 1. Moreover, after normalization, all the variables are normally distributed, and the impact of each factor in the regression models on transfer ridership can then be compared. Therefore, when comparing the coefficients, the variables have been normalized into the same scale.”

(2) Please check Lines 565-569 on Page 20 of the resubmitted manuscript.

“For the metro-to-bus mode, it can be seen that the effect of transfer time on transfer ridership is the most significant among all independent variables, about 1.22 to 35.45 times that of other variables. Those numbers are derived based on the estimated coefficients: 2.127/0.06 = 35.45. In this study, we compare the impact of each factor on the transfer ridership by the estimated influencing coefficients.”

We hope that these revisions in the resubmitted manuscript can meet your expectations.

Comment 8:

8.     The paper developed separate ridership models for time during workdays + weekends and during holidays. The author should explain why it is conducted in this way, and why combining workdays and weekends together is a better option than combining weekends and holidays in one group or separating them into three groups.

Response:

Thanks for your comments. This paper developed separate ridership models for the time during workdays + weekends and holidays. We also explain why it is conducted in this way, and why combining workdays and weekends is a better option than combining weekends and holidays in one group or separating them into three groups in the resubmitted manuscript. This is because the distribution of metro ridership, transfer ridership, and bus ridership is very similar on weekends and weekdays. The holiday distribution is significantly different from the weekday distribution. Figures 6, 7, and 8 just confirm this. Please check Lines 343-353 on Pages 11-12, Lines 365-374 on Pages 12-13, and Lines 390-410 on Pages 13-14 of the resubmitted manuscript.

“Figure 6. (a1), (b1), and (c1), show that there is a strong consistency in the distribution of metro ridership, inbound ridership, and outbound ridership from 6:00 to 22:59 on weekdays. Moreover, the distribution of metro ridership, inbound ridership, and outbound ridership shows the same significant morning peak and evening peak. The morning peak hours are from 7:00-9:00. And the evening peak hours are from 17:00-20:00. The distribution of metro ridership, inbound ridership, and outbound ridership has a strong consistency on Saturday with significant morning and evening peak hours. Peak hours on Saturdays are similar to that on weekdays.

From Figure 7, it can be seen that the distribution of transfer ridership is similar to the distribution of metro ridership. The distribution of transfer ridership on weekdays has a strong consistency with a significant morning and evening peak. The morning peak hours are from 7:00-9:00. And the evening peak hours are from 17:00-19:00. For metro-to-bus transfer mode, the transfer ridership in the evening peak is higher than that in the morning peak, while the opposite is true for the transfer ridership of the bus-to-metro mode. The distribution of transfer passenger flows on Saturdays is similar to that of weekdays, which also have significant morning and evening peaks. This may be because some commuters also work on weekends.

From Figure 8, it can be seen that the distribution of bus ridership on weekdays has a strong consistency, and the bus ridership at the same moment on different days is almost the same with a significant morning and evening peak. Similarly, the distribution of bus ridership each Saturday is also very consistent and has significant morning and evening peaks. The morning peak hours are from 7:00-9:00. The evening peak hours are from 17:00-19:00.

In short, transfer ridership, transfer time, bus ridership, and metro ridership are distributed in different patterns at different times. It is necessary and important to classify these factors by different types of dates for exploring the determinants of transfer ridership. Therefore, this paper developed separate ridership models for the time during workdays + weekends and holidays.”

We hope that these revisions in the resubmitted manuscript can meet your expectations.

Comment 9:

9.     Line 539 “High temperatures, heavy rain, or low visibility conditions can increase transfer ridership”, and line 553 “under the conditions of high temperature, heavy rainfall, strong wind, or low visibility, more travelers choose to transfer from the bus to the metro”.  Those results are somewhat counter-intuitive. Any explanation for this result? Also, some variables are not significant at all, for example rainfall in metro-to-bus mode has a p-value of 0.879, which is not significant enough to support the statement that heavy rainfall increase transfer ridership”.

Response:

Thank you for your comments. Lines 613-614, “High temperatures, heavy rain, or low visibility conditions can increase transfer ridership.” This result is derived from the effect of each factor on the transfer ridership for the metro-to-bus mode during holiday periods in Table 4. And lines 626-627 “under the conditions of high temperature, heavy rainfall, strong wind, or low visibility, more travelers choose to transfer from the bus to the metro”. This result is derived from the effect of each factor on the transfer ridership for the bus-to-metro mode during holiday periods in Table 4. This is because usually transfer passengers cannot take a single vehicle directly to their destination and have to transfer. On holidays, people's travel time is more flexible. Moreover, in bad weather, such as high temperatures, heavy rain, strong wind, or low visibility conditions, people prefer to take public transportation over other transfer modes, such as shared bikes, electric bikes, or walking. Therefore, bad weather will increase the transfer ridership. In other words, in bad weather, for those transfer passengers who took the metro or bus for the first leg of their journey and they had to make a transfer to get to their destination, most of them will choose to take the bus or metro to reach their destination on the second leg of their journey. As a result, bad weather will increase the transfer ridership. Please check Lines 634-646 on Pages 21-22 of the resubmitted manuscript.

“Furthermore, bad weather will increase the transfer ridership. This is because usually transfer passengers cannot take a single vehicle directly to their destination and have to transfer. On holidays, people's travel time is more flexible. Moreover, in bad weather, such as high temperatures, heavy rain, strong wind, or low visibility conditions, people prefer to take public transportation over other transfer modes, such as shared bikes, electric bikes, or walking. Therefore, bad weather will increase the transfer ridership. In other words, in bad weather, for those transfer passengers who took the metro or bus for the first leg of their journey and they had to make a transfer to get to their destination, most of them will choose to take the bus or metro to reach their destination on the second leg of their journey. As a result, high temperatures, heavy rain, or low visibility conditions can increase transfer ridership of the metro-to-bus mode, and under the conditions of high temperature, heavy rainfall, strong wind, or low visibility, more travelers choose to transfer from the bus to the metro. ”

Also, some variables are not significant at all, for example, rainfall in metro-to-bus mode has a p-value of 0.879. But the result is from Table 3 on line 555, which is the result of the metro-to-bus mode on workdays and weekends, not during holidays. This does not correlate with the two findings above. Please check Line 555 on Pages 20-21 of the resubmitted manuscript.

We hope these revisions can answer your question.

Comment 10:

10.  What is the performance or goodness of fit for the developed models?

Response:

Thanks for your comments. It is necessary to provide the model fit statistics to justify the final specification of the model. Therefore, in Table 3, Table 4, and Table 5, we added three diagnostic statistics of models, namely null deviance, residual deviance, Akaike information criterion (AIC) and . Please check Line 560 on Page 19, Line 653 on Page 22, Line 674 on Page 23, and Lines 533-538 on Page 18 of the resubmitted manuscript.

“To show how well the GPR model fit the data, we used  to evaluate the goodness of fit of the model [69], the formula is as follows:

Where n is the sample size of the dependent variable,  is the real values of the dependent variable,  is the mean value of the real dependent variable, and  is the fitted values of the dependent variable by the GPR model.”

We hope these revisions can answer your question.

Comment 11:

11.  In the conclusions, the author concluded “It is feasible to adopt the GPR model to investigate the influence of each factor on transfer ridership of different transfer modes on weekdays, holidays, and typhoon weather, respectively.” Without showing how well the GPR model fit the data, this conclusion is not supported by the study.

Response:

Thanks for your comments. We added the goodness-of-fit of the GPR models in the resubmitted manuscript. It shows how well the GPR model fits the data. In the conclusions, we concluded, “It is feasible to adopt the GPR model to investigate the influence of each factor on transfer ridership of different transfer modes on weekdays, holidays, and typhoon weather, respectively.” This conclusion is supported by the study. Please check Line 560 on Page 19, Line 653 on Page 22, Line 674 on Page 23, and Lines 533-538 on Page 18 of the resubmitted manuscript.

“To show how well the GPR model fit the data, we used  to evaluate the goodness of fit of the model [69], the formula is as follows:

Where n is the sample size of the dependent variable,  is the real values of the dependent variable,  is the mean value of the real dependent variable, and  is the fitted values of the dependent variable by the GPR model.”

We hope that these revisions in the resubmitted manuscript can meet your expectations.

Comment 12:

12.  Line 730, “other non-linear models should be used to discuss the relationship between any existing nonlinear”, there are some approaches to capture possible non-linear relationships between dependent and independent variables in the developed models, for example adding polynomial terms for the independent variables.

Response:

Thanks for your suggestions. Line 730, “other non-linear models should be used to discuss the relationship between any existing nonlinear”, there are some approaches to capture the possible non-linear relationship between dependent and independent variables in the developed models, for example adding polynomial terms for the independent variables. In the subsequent study, we will explore the possible non-linear relationship between dependent and independent variables by adding polynomial terms for the independent variables. Please check Lines 890-894 on Page 28 of the resubmitted manuscript:

“Moreover, in this study, other nonlinear models should be used to discuss the relationship between any existing nonlinearities. In the subsequent study, we will explore the possible non-linear relationship between dependent and independent variables by adding poly-nomial terms for the independent variables.”

We hope that these revisions in the resubmitted manuscript can meet your expectations.

Comment 13:

13. Besides revealing the relationship between transfer ridership and various critical factors, is there any other contribution or benefits like where those developed model can be applied in real life and how the results of this paper can help decision makers in term of improving the transit planning etc.

Response:

Thanks for your comments. Besides revealing the relationship between transfer ridership and various critical factors, there are also other benefits like where those developed models can be applied in real life. And the results of this paper can help decision-makers in terms of improving transit planning etc. Please check Lines 823-840 on Page 27 of the resubmitted manuscript:

“Besides revealing the relationship between transfer ridership and various critical factors, there are also other benefits, the GPR model can be applied in real life. The GPR model can be also used to analyze the impact of various factors on metro ridership or bus ridership. Moreover, the results of this paper can help decision-makers in terms of improving transit planning, etc. For example, bad weather can promote an increase in transfer ridership on holidays, the public transport practitioners should ensure the safe travel of transfer passengers and establish effective measures to defend against bad weather. In typhoon weather, transfer ridership is affected by the weather, so public transportation management should focus on safe transfers of the transfer passengers in the extreme weather. The economic level near the metro stations has a negative impact on the transfer ridership. Therefore, in regions with poor economic levels, transportation decision-makers should improve the accessibility of bus and subway networks and improve the transfer service. In areas with high pedestrian density, which usually have a great transfer ridership, transportation decision-makers should improve transfer efficiency at these stations, such as reducing bus headways and adjusting metro schedules. In brief, these important findings can be used in real life and can help transportation decision makers and managers to improve public transportation network planning, and improve metro and bus stations to enhance the service capacity and attractiveness of public transportation.”

We hope that these revisions in the manuscript can meet your expectations, and these explanations can solve your question about the results.