Determinants of Bicycle Transfer Demand
at Metro Stations
Analysis of Stations in Nanjing, China
Lijun Chen, Adam J. Pel, Xuewu Chen, Daniel Sparing, and Ingo A. Hansen
Metro transit networks are constantly expanding to meet the growing
travel demand that accompanies rapid urban growth. One main dis-
advantage of metro transport is its heavy reliance on access and egress
transport and, hence, on the corresponding transfer facilities. Improve-
ments to these transfer facilities therefore have the potential to increase
use of the metro system while possibly alleviating the traffic load on the
urban road network. Because bicycle transfer at metro stations—that is,
using a bicycle as mode of access to or egress from the metro system—
is underused, the determinants of the demand for bicycle transfer are
investigated. Results and findings are valuable for designing policies
aimed at improving metro ridership and for designing bicycle parking
and transfer facilities at metro stations. To this end, several metro stations
in Nanjing, China, were analyzed, and two stereotypical metro stations
were selected for how well they represented the system. A large-scale
survey was conducted on metro travelers’ opinions on and use of bicycle
transfer facilities, and data were collected on the current attributes of
service groups, bicycle parking occupancy, and transfer mode alterna-
tives. Furthermore, metro travelers’ (latent) transfer preferences for
bicycle rental facilities were investigated. Two transfer choice models
were estimated to identify and quantify the determinants for bicycle
transfer demand: one focuses on current walk–metro trips, and the
other focuses on current bus–metro trips. The explanatory determinants
are discussed, and relative weights are computed with multiple linear
regression analysis.
The metro system is the most promising public transport mode for
alleviating the traffic load on the urban road infrastructure. Meanwhile,
rapid urban growth and accompanying increases in travel demand
necessitate a constant expansion of the metro transit network. At
present, more than 20 of the larger Chinese cities have developed
or are developing an urban rail system (1). One of the main dis­
advantages of metro transport is that it heavily relies on access and
egress transport and, therefore, on the corresponding transfer facili-
ties [because up to 42% of total travel time typically is spent on
L. Chen, A. J. Pel, D. Sparing, and I. A. Hansen, Faculty of Civil Engineering and
Geosciences, Department of Transport and Planning, Delft University of Technology,
Stevinweg 1, P.O. Box 5048, 2628 CN Delft, Netherlands. Current affiliation for
L. Chen: Faculty of Transportation, Transport and Planning Department, Southeast
University, No. 2, Sipailou, 210096 Nanjing, China. X. Chen, Faculty of Transporta-
tion, Transport and Planning Department, No. 2, Sipailou, 210096 Nanjing, China.
Corresponding author: L. Chen, chenlijun2011@hotmail.com.
Transportation Research Record: Journal of the Transportation Research Board,
No. 2276, Transportation Research Board of the National Academies, Washington,
D.C., 2012, pp. 131–137.
DOI: 10.3141/2276-16
131
access and egress trips (2)], and transferring generally is regarded as
stressful and time-consuming (3). Currently, about 65.5% of metro
users must transfer at least once during an access or egress trip, and
an average of 0.9 transfer is made within one metro trip (4). This
fact clearly argues in favor of improving transfer facilities, thereby
potentially increasing metro system usage and alleviating the traffic
load on the urban road network.
Several studies have been conducted on transfer services with
respect to urban rail transit; most of this research has focused on
improving intermodal transfer for bus–rail transport. For example,
Brown and Thompson examine the relationship among service
orientation, bus–rail service integration, and transit performance
in several U.S. metropolitan areas (5). It is suggested that transit
managers must focus on the relationship between service orientation
and bus–rail integration to serve customers better and improve over-
all transit performance. Seaborn et al. consider bus–metro transfer
combinations and make recommendations on the maximum elapsed
time thresholds to identify transfers between journey stages for pas-
sengers on the London Underground network (6). Dickins argues
that the fast-paced motorization of city centers leads to large-scale
traffic problems, particularly in the morning and evening peaks, and
that urban metro transit would benefit from adequate park-and-ride
facilities (7).
Even though much research has been done on automobile and
public transport transfer facilities at metro stations, bicycle transfer
facilities at metro stations have received less attention from urban
transport policy makers and planners. Most of the current literature
in this area comes from countries where cycling is a main mode
of transportation. Martens’s comparative analysis of bike-and-
ride patterns in the Netherlands, Germany, and the United Kingdom
finds strong similarities in the characteristics of bike-and-ride trips
and users, despite the different bicycle cultures and infrastructure
facilities in these countries (8). The small variations in bicycle use
that do exist are explained in terms of differences in altitude, city
size, population features, ethnic composition, cultural conditions,
and so on (9). Moskovitz and Wheeler argue that for a proper analysis
of bicycle usage and bicycle–transit transfer, bicycle parking facilities
must be considered; they propose a method for collecting bicycle
parking data (10). Finally, Tang et al. study the influence of bicycle
parking and transfer facilities on bicycle usage in the Chinese cities
of Beijing, Shanghai, and Hangzhou and find that only a small share
of rail transit riders currently choose the bicycle as their access or
egress mode (11). The choice is related to the distances between a
user’s residential location, the preferred bus stop, and the preferred
rail station as well as several bus service characteristics.
Low bicycle usage typically is ascribed to poor bicycle infra-
structure (12) or insufficient bicycle parking facilities combined
132 Transportation Research Record 2276

with travelers’ concerns regarding safety (and the possibility of

bicycle theft) (13). Meanwhile, common reasoning is that improv-
ing bicycle parking and transfer services at metro stations will Yangtze River
encourage more travelers to choose the bicycle as their preferred
access or egress mode (2). This reasoning is supported by an
Line 1
observed increase in bicycle access to metro in the Netherlands after
a national program to upgrade and better secure bicycle parking
facilities at train stations (14, 15). Therefore, it is essential to identify
what determines better bicycle transfer facilities.
In this paper, the determinants of bicycle transfer demand at metro
stations are investigated in an effort to improve the transfer service
level. The results and findings presented are thus valuable for design-
ing policies aimed at improving metro ridership as well as bicycle
parking and transfer facilities at metro stations. To this end, several
metro stations in Nanjing, China, were analyzed, and two stereo-
typical metro stations were selected for their representativeness of
the system. A large-scale survey was conducted on metro travelers’
opinions on and usage of bicycle transfer facilities, and data were
collected regarding the current attributes of service groups, bicycle
parking occupancy, and transfer mode alternatives.
The rest of the paper is organized as follows. Metro travelers’ Line 2
(latent) transfer preferences for bicycle rental facilities are inves- Shengtai Road Station
tigated. Then, the characteristics of bicycle transfer at the two Longmian Road Station

targeted Nanjing metro stations are discussed. Next, two transfer

choice models are discussed that are estimated to identify and quan-
tify the determinants for bicycle transfer demand: one focuses on
current walk–metro trips, the other on current bus–metro trips. The
explanatory determinants are discussed, and relative attribute levels
are computed with the use of multiple linear regression analysis. In
the final section, concluding remarks are presented regarding the
main findings and practical implications for the design of bicycle
transfer facilities.
Characteristics of
Bicycle–Metro Transfer
First, the processes of collecting data and creating the metro traveler
questionnaire are discussed. Then, the most interesting observations
derived from the collected data set are presented.
Data Collection and Traveler Questionnaire
The survey data used in this study come from an extensive question-
naire investigation undertaken in April and May 2011 in Nanjing,
China. Nanjing is located at the Yangtze River Delta in eastern
China. It is one of the country’s most important cities and serves as
a national hub for education, research, transportation, and tourism.
It is the second-largest commercial center in eastern China, after
Shanghai. Two metro lines with a total of 57 stations operate in the
city, presently providing 585,000 passenger trips per day, with the
metro passenger flow continuously increasing (Figure 1).
For the following analysis, the focus is on two stereotypical metro
stations: Longmian Road and Shengtai Road. Longmian Road Sta-
tion is located in a typical dwelling district, whereas Shengtai Road
Station is located in a typical shopping and business district with
industry, universities, and shopping facilities nearby (Figure 2).
These two stations were selected as together being representative
of the Nanjing metro transit system. Correspondingly, the question-
FIGURE 1   Metro network in Nanjing, China (Longmian Road
and Shengtai Road Stations indicated in red).
naire responses can be considered to capture adequately the various
transfer determinants and behavior observed throughout the entire
metro transit system.
To understand characteristics of travelers and their preferences
for bicycle–metro transfer, a survey was conducted that yielded 500
responses at these two metro stations and another 1,284 responses in
the surrounding neighborhoods. The following data were collected:
• Metro transit trip characteristics (e.g., travel purpose, travel time,
and travel distance),
• Transfer characteristics (e.g., access transfer mode, egress
transfer mode, access travel time, and egress travel time),
• Demographic characteristics (e.g., gender, age, bicycle owner-
ship, and income level),
• Familiarity characteristics (e.g., attitude toward bicycle rental,
free time preference, and focus of attention), and
• Scenario preferences (e.g., bus–metro, bicycle–metro, or walk–
metro transfer mode).
Characteristics of Bicycle Transfer
Transfer Mode Shares
Chen and Chen apply transfer priority sequences to different types
of urban rail stations and find that for stations in dwelling districts,
walking is the most common access mode, followed by bicycle, then
bus; for stations in shopping districts, walking is the most common
access mode, followed by bus, then bicycle (16).
Chen, Pel, Chen, Sparing, and Hansen 133

(a) (b)

FIGURE 2   Satellite images of (a) Longmian Road Station and (b) Shengtai Road Station [metro station entrances and exits indicated
in yellow, and nearby locations (e.g., supermarkets, shopping malls, restaurants) indicated in white].

The survey data collected in the present study show similar findings Bicycle Versus Walking as Access or Egress Mode
(Table 1). In general, cycling is widely accepted by metro travelers
as an adequate access or egress mode, but the bicycle access and The typical travelers who choose bicycle transfer otherwise would
egress shares are rather different for the two stations. At Longmian have to accept walk long distances or wait a long time for a bus (9).
Road Station (dwelling district), bicycles are more often used for the Acceptable walking times vary, and responses to the questionnaire
access trip; at Shengtai Road Station (shopping district), bicycles are indicate that the share of metro travelers switching from walking to
more often used for the egress trip. In addition, the bicycle share is cycling as the preferred mode of access or egress depends on walk-
observed as a more preferred mode of egress than in other studies (2). ing time (Figure 4). For instance, when walking time would exceed
This difference can be explained by the fact that a reasonably large 15 minutes, approximately 65% of respondents prefer cycling.
share of survey respondents work in the vicinity of the Shengtai
Road Station and prefer cycling as their mode of egress transport.
The authors also noted while conducting the survey that several Bicycle Parking Time and Duration
respondents used folding bicycles for their access or egress trip at Bicycle–Metro Transfer Facilities
(however, this information was not included in the survey).
Moskovitz and Wheeler categorize bicycle parking durations as
short-term (less than 2 hours), midterm (2 to 4 hours), and long-term
Travel Purposes for Bicycle–Metro Trips (more than 4 hours) (10). In this study, the parking durations that were
observed are respectively less than 2 hours, between 2 and 4 hours,
The use of bicycle as a mode of transport for access, egress, or both between 4 and 6 hours, between 6 and 8 hours, and more than 8 hours
is strongly related to the purpose of the metro trip. Figure 3 illustrates (Figure 5). Most bicycles are parked at bicycle–metro transfer
the trip purpose shares for all observed bicycle–metro trips. Travelers facilities for 4 to 6 hours at Shengtai Road Station and more than
effectively never use a bicycle as the preferred access or egress transfer 8 hours at Longmian Road Station. Because Shengtai Road Station is
mode for trip purposes that are highly time-sensitive, such as business in a shopping district, more bicycles (in total) are parked for shorter
trips (1%); they tend to choose cycling often for trip purposes that are
intermediately time-sensitive, such as school (12%) and work (22%),
and most often choose cycling for non-time-sensitive trip purposes Work
such as visiting friends (22%) and shopping (27%). 30.00%
25.00%
Other 20.00% Business
TABLE 1   Passenger Transfer Modes at Metro Stations 15.00%
10.00%
Longmian Road Station Shengtai Road Station 5.00%
0.00%
Mode Access (%) Egress (%) Access (%) Egress (%)
Entertainment School
Bicycle 12.4 5.4 5.8 8.5
Bus 26.7 36.8 34.7 38.4
Walking 43.4 42.6 31.7 39.5
Other 17.5 15.2 21.7 7.4 Visiting friends Shopping
Total 100 100 100 100
FIGURE 3   Travel purposes for bicycle–metro trips.
134 Transportation Research Record 2276

100.00%
90.00%
80.00%

Share of Bicycle Usage

70.00%
60.00%
50.00%
40.00%
30.00%
20.00%
10.00%
0.00%
5 mins 10 mins 15 mins 20 mins 25 mins 30 mins
Critical Walking Time

FIGURE 4   Bicycle as preferred mode of access or egress depending on walking time.

times; because Longmian Road Station is in a typical dwelling district,
fewer bicycles (in total) are parked for longer times.
Different station types also exhibit different bicycle parking time
patterns (Figure 6). At Longmian Road Station, most bicycles are
parked in the morning, possibly related to travelers’ work, school, or
shopping trips; by contrast, at Shengtai Road Station, most bicycles
are parked in late morning and early afternoon. These observations
indicate that the type of station (i.e., land use of the surrounding dis-
trict) has a large influence on the usage of bicycle parking facilities
and thus must be considered in the facility design process.
Familiarity and Preferences Regarding
Bicycle Rental Facilities
Bicycle rental facilities can potentially increase bicycle transfer
shares, but research in this area is limited. Survey results indicate
that, unfortunately, only 4.6% of respondents were familiar with and
used these facilities; investigation clearly is warranted. Meanwhile,
40.0% of respondents were not familiar with the facilities but would
be interested in using such services; 29.2% were familiar with but
not interested in using such services, and 25.8% were unfamiliar
and uninterested.
Even though most travelers would not prefer bicycle rental
facilities, results exhibit a large potential to increase rental bicycle
usage for access or egress trips. The survey inquired as to how trav-
elers were now informed about bicycle rental facilities. Responses
indicated that the main source of relevant information was television
commercials (34%), followed by newspapers (19%) and signaling
at the metro stations (19%); Internet (16%); and finally friends,
colleagues, and other acquaintances (10%). Simply finding bicycle
rental facilities at a station was mentioned as a preferred way of
being informed. The visibility of these facilities clearly remains to
be improved, because at present most people (approximately 70%)

(b)
(a)

FIGURE 5   Bicycle parking duration at (a) Longmian Road Station and (b) Shengtai Road Station.
Chen, Pel, Chen, Sparing, and Hansen 135

600

Number of Bicycles Parking

500

400

300

200

100

0
0

0
:0

:0

:0

:0

:0

:0

:0
-9

1
-1

-1

-1

-1

-1

-2
00

00

0
7:

:0

:0

:0

:0

:0
9:

11

13

15

17

19
Time of Day

Longmian Road Station Shengtai Road Station

FIGURE 6   Bicycle parking time.

are informed by the media and a minority (approximately 30%) are
informed via personal experiences or acquaintances.
Travelers’ main concerns regarding bicycle rental differ. The two
most important reasons for choosing to rent a bicycle for access or
egress trips are rental convenience (29%) and rental location (24%);
the rental fee per trip comes next (10%). A minority of respondents
indicated being concerned about bicycle quality (8%), a rental card
fee (4%), and bicycle management (3%). Responses strongly suggest
improving the convenience and locations of rental facilities.
To promote the usage of rental bicycles for access and egress trips,
a free usage period often is used. A metro passenger can use a bicycle
for access and egress trips free for the first x hours and pay only if
the bicycle is returned to any rental location at a metro station after
the specified period. The survey investigated the amount of free
usage time that respondents think should be sufficient. Most respon-
dents (57%) considered 2 hours of free time appropriate, whereas 33%
would prefer such facilities if 3 hours of free time were provided; only
10% expected that 1 hour would be sufficient.
Bicycle–Metro Transfer: Determinants
In this section, the collected survey data on the characteristics
of bicycle transfer presented in the previous section are used to
estimate two transfer choice models and to identify and quantify
the determinants for bicycle transfer demand. First, the applied
choice model is introduced, then the explanatory determinants
are discussed.
walking as their alternative access or egress mode, and the other con-
siders travelers with bus as their alternative mode. The reason for this
separation is that the determinants for these two transfer choices are
very different.
The transfer choice model describing walking versus cycling
as the preferred access or egress mode considers the following set
of explanatory variables: age, bicycle ownership, gender, income,
cycling distance from the traveler’s trip origin to the bicycle parking
facility (OR-BPF), and walking distance between the bicycle parking
facility and the metro station (BPF-MS). The dependent variables are
the transfer alternatives walk–metro and bicycle–metro.
The transfer choice model describing bus versus bicycle as the pre-
ferred access or egress mode considers the following set of explanatory
variables: age, bicycle ownership, gender, income, cycling distance
between the traveler’s trip origin and the bicycle parking facility
(OR-BPF), walking distance between the bicycle parking facilities
and the metro station (BPF-MS), walking distance between the trav-
eler’s trip origin and the bus stop (OR-BUS), cost of a bus ticket, and
waiting time at the bus stop. The dependent variables are the transfer
alternatives bus–metro and bicycle–metro.
Both models were estimated with the collected survey data and
a multiple linear regression analysis. The regression outcomes are
presented next.
Bicycle Transfer Determinants
Walking Versus Bicycle Transfer Determinants

In the first choice model, travelers’ choice between walking and

Transfer Choice Model
To identify the bicycle transfer determinants, a utility-based transfer
choice model was developed. Expected utility maximization was used
because this theory is well suited for describing travelers’ rational and
informed choice behavior between independent choice alternatives.
Two choice models were developed: one considers travelers with
bicycle as their preferred access or egress mode is considered. The
explanatory variables that were found not to be significant are age,
bicycle ownership, gender, and income. These findings are in line with
early findings by Krygsman et al., who find that sociodemographic
characteristics play no significant role in mode choice for transport
access or egress (17). Parameter values for the significant coefficients
are listed in Table 2. The estimated model coefficients show a default
136 Transportation Research Record 2276

TABLE 2   Estimated Coefficients for Bicycle–Metro Transfer, Concluding Remarks

with Walking as Alternative
In this paper, the determinants of bicycle transfer demand at metro
Coefficients stations were investigated. A large-scale survey was conducted at
Unstandardized Standardized two Nanjing metro stations regarding metro travelers’ preferences and
usage of bicycle transfer facilities. Several characteristics of bicycle
Parameter B SE Beta t-Value Sig. transfer were discussed, two transfer choice models were estimated,
and travelers’ determinants for bicycle transfer demand were identified
Constant −.194 .033 ­— −5.975 .000 and quantified.
OR-BPF (m)  .001 .000  .702 21.042 .000 The main findings presented in this paper have several direct
BPF-MS (m)  .000 .000 −.071 −2.142 .033 practical implications. First, survey results indicate that more than
one-half of metro users prefer to use bicycle transfer services. This
Note: SE = standard error; sig. = significance; — = not applicable.
outlook is promising for policy and station design planners who aim
to increase bicycle transfer usage. The main service groups for such
facilities are travelers with more than 15 minutes walking distance
modal preference for walking (negative constant), whereas travelers from the metro station and travelers making non-time-sensitive trips
will gradually switch to cycling in case of longer distances from (e.g., for shopping, visiting friends, and—to a lesser extent—going
origin to bicycle parking facilities (OR-BPF) (which may operate to work and school).
as a proxy variable for the alternative walking distance and explains Second, the land use of the surrounding neighborhoods affects how
the positive coefficient value). The distance from bicycle parking bicycle transfer facilities are used and therefore should be considered
facilities to the metro system (BPF-MS) has a negative impact on while designing such facilities. The results presented here suggest
the choice to cycle, a finding to be expected because a more distant that a bicycle transfer facility at a station in a shopping district may
facility makes the bicycle alternative less attractive. Nevertheless, serve more travelers for shorter parking times, whereas the same
the relative weight of the former travel distance is much higher than facility at a station in a dwelling district may serve fewer travelers
that of the latter. The R2 of the estimated model is .447. for longer parking times. Also, bicycle transfer facilities have higher
potential usage at stations in dwelling districts than in shopping
districts.
Bus Versus Bicycle Transfer Determinants Third, bicycle rental services have good potential to stimulate
bicycle transfer at metro stations. For this potential to be realized,
In the second choice model, travelers’ choice between bus and
the focus should be on familiarizing travelers with the bicycle rental
bicycle as their preferred mode of access or egress is considered.
options and improving the visibility of the bicycle rental facilities.
The explanatory variables that were found not to be significant are age,
Because travelers’ two most important concerns related to choosing
bicycle ownership, gender, BPF-MS, and OR-BUS. These findings
rental bicycles are convenience and location, these issues should be
are in line with early findings by Krygsman et al., who find that socio­
considered when new rental services are planned.
demographic characteristics play no significant role in mode choice
for transport access or egress (17). Parameter values for the signifi- Finally, travelers generally prefer walking as the access or egress
cant coefficients are shown in Table 3. The estimated model coeffi- mode but will choose bicycle transfer when the distance between
cients show a default modal preference for cycling (positive constant), the origin and the bicycle parking facility is long and the bicycle
whereas travelers gradually choose bus travel in the case of larger parking facility is close to the metro system. Travelers may choose
travel distances (OR-BPF), higher income, lower bus ticket fares, bus over bicycle if the distance between the origin and the bicycle
or lower waiting time at bus stops. The R2 of the estimated model parking facility is too long, the bus fare is affordable (for users with
is .470. high income levels), and waiting times at the bus stop are short. The
estimated model parameters lead to a reasonably good fit (R2 = .447
and .470, respectively). Therefore, future research is recommended
on how these findings and practical implications can be translated
TABLE 3   Estimated Coefficients for Bicycle–Metro Transfer, into practice in studies of metro station layout and bicycle transfer
with Bus as Alternative
facility design.
Coefficients

Unstandardized Standardized Acknowledgments

Parameter B SE Beta t-Value Sig. The authors thank the students who were involved in the data
collection process at Southeast University, Nanjing, China.
Constant .578 .087 — 6.652 .000
OR-BPF (m) .000 .000 −.707 −17.098 .000
Income (yuan) −.117 .022 −.199 −5.405 .000 References
Cost of bus .164 .027 .278 6.013 .000
ticket (yuan)   1. Shi, Z. Planning for the Construction of Urban Rail Transit (in Chinese).
Waiting time .008 .003 .144 3.377 .001 Urban Rapid Rail Transit, Vol. 2, 2004, pp. 25–33.
(min)   2. Pan, H., Q. Shen, and X. Song. Intermodal Transfer Between Bicycles
and Rail Transit in Shanghai, China. In Transportation Research Record:
Note: — = not applicable. Journal of the Transportation Research Board, No. 2144, Transportation
Chen, Pel, Chen, Sparing, and Hansen
Research Board of the National Academies, Washington, D.C., 2010,
pp. 181–188.
  3. Iseki, H., and B. Taylor. Not All Transfers are Created Equal: Towards
a Framework Relating Transfer Connectivity to Travel Behavior.
Transport Reviews, Vol. 29, No. 6, 2009, pp. 777–800.
  4. Li, D., and K. Yan. Optimization Technology of the Configuration of Joint
Transportation Resources at Urban Rail Passenger Station (in Chinese).
PhD dissertation. Tongji University, Shanghai, China, 2010.
  5. Brown, J. R., and G. L. Thompson. Service Orientation, Bus–Rail Service
Integration, and Transit Performance: Examination of 45 U.S. Metro­
politan Areas. In Transportation Research Record: Journal of the Trans-
portation Research Board, No. 2042, Transportation Research Board of
the National Academies, Washington, D.C., 2008, pp. 82–89.
  6. Seaborn, C., J. Attanucci, and N. H. M. Wilson. Analyzing Multimodal
Public Transport Journeys in London with Smart Card Fare Payment
Data. In Transportation Research Record: Journal of the Transporta-
tion Research Board, No. 2121, Transportation Research Board of the
National Academies, Washington, D.C., 2009, pp. 55–62.
  7. Dickins, I. S. J. Park and Ride Facilities on Light Rail Transit Systems.
Transportation (Netherlands), Vol. 18, No. 1, 1991, pp. 23–36.
  8. Martens, K. The Bicycle as a Feedering Mode: Experiences from Three
European Countries. Transportation Research Part D, Vol. 9, No. 4,
2004, pp. 281–294.
  9. Rietveld, P., and V. Daniel. Determinants of Bicycle Use: Do Municipal
Policies Matter? Transportation Research Part A, Vol. 38, No. 7, 2004,
pp. 531–550.
10. Moskovitz, D. A., and N. Wheeler. Bicycle Parking Analysis with Time
Series Photography. In Transportation Research Record: Journal of the
137
Transportation Research Board, No. 2247, Transportation Research
Board of the National Academies, Washington, D.C., 2011, pp. 64–71.
11. Tang, Y., H. Pan, and Q. Shen. Bike-Sharing Systems in Beijing,
Shanghai, and Hangzhou and Their Impact on Travel Behavior. Pre-
sented at 90th Annual Meeting of the Transportation Research Board,
Washington, D.C., 2011.
12. Design Manual for Bicycle Traffic. CROW, Ede, Netherlands, 2007.
13. Shaheen, S. A., H. Zhang, E. Martin, and S. Guzman. China’s Hangzhou
Public Bicycle: Understanding Early Adoption and Behavioral Response
to Bikesharing. In Transportation Research Record: Journal of the
Transportation Research Board, No. 2247, Transportation Research
Board of the National Academies, Washington, D.C., 2011, pp. 33–41.
14. Martens, K. Promoting Bike-and-Ride: The Dutch Experience. Trans-
portation Research Part A, Vol. 41, No. 4, 2007, pp. 326–338.
15. Rietveld, P. The Accessibility of Railway Stations: The Role of the
Bicycle in the Netherlands. Transportation Research Part D, Vol. 5,
No. 1, 2000, pp. 71–75.
16. Chen, L., and X. Chen. Transferring Characteristic of Urban Rail Stations
Based on Stations Functional Orientation. Proc., 10th International
Conference of Chinese Transportation Professionals: Integrated
Transportation Systems—Green and Intelligent and Reliable, 2010,
pp. 145–154.
17. Krygsman, S., M. Dijst, and T. Arentze. Multimodal Public Transport:
An Analysis of Travel Time Elements and the Interconnectivity Ratio.
Transport Policy, Vol. 11, No. 3, 2004, pp. 265–275.
The Intermodal Transfer Facilities Committee peer-reviewed this paper.