Sunteți pe pagina 1din 17

Journal of Transport & Health 10 (2018) 236–252

Contents lists available at ScienceDirect

Journal of Transport & Health


journal homepage: www.elsevier.com/locate/jth

Choosing safe and suitable bicycle routes to integrate cycling and


T
public transport systems
⁎ ⁎
M. Saplıoğlua, , M.M. Aydınb,
a
Department of Civil Engineering, Suleyman Demirel University, Isparta, Turkey
b
Department of Civil Engineering, Gumushane University, Gumushane, Turkey

A R T IC LE I N F O ABS TRA CT

Keywords: Bicycle utilization has become an important transportation mode in various cities during the past
Bicycle-public transport integration decade. It has brought a rapid growth as an important transport mode in many countries by
Analytic hierarchy process providing several benefits for traffic, health and cost. However, the growing in bicycle utilization
Geographic information systems demand causes many problems such as safety, effectiveness etc. For this reason, other effective
Survey
parameters on urban transport should be considered when planning new safe bicycle paths in the
spatial platform. In this study, it was aimed to examine all these negative and positive effective
parameters on safety and choice while integrating cycling with a public transport (bus link)
system. To determine effective parameters on route choice, a Questionnaire Survey (QS) was
carried out on 460 participants who are cycling for a long time in Isparta City/Turkey. It was
obtained from the QS that Accident Prone Areas (APA) are the most important factor on cycling
and public transport integration. The other parameters are determined as Bus Lane (BL), Road
Side Car Park (RSCP), Bicycle Parks (BP), Road Grade (RG), Signalization (S), Traffic Capacity
(TC), Connected Bike Lane (CBL) and Separated Bike Lane (SBL) according to their importance.
In order to determine safer and serviceable bicycle routes selection problem, accident-prone
locations (using Accident Reports and Geographic Information Systems-GIS), bus lines and survey
data has been used in GIS. Also, a multi-criteria decision making approach Analytic Hierarchy
Process (AHP) method is implemented to analyze survey data and decide effective parameters on
safer and serviceable bicycle routes. It was found that safer and serviceable bicycle routes can be
determined by the utilization of QS, GIS and AHP methods with together. Also, it was concluded
that Accident Prone Areas, Bus Lanes and Road Side Car Parks are the most important factors
while integrating cycling with a public transport system.

1. Introduction

The integration of transportation modes which coordinates harmony of all transport modes such as pedestrian, bicycle, motor
vehicle, bus and railway systems, is an important issue for safety and sustainability of urban transportation systems. Also, bicycle and
public transport co-existence is a particularly necessary element for sustainable urban transport. But while these systems are oper-
ating on the same route, different safety and performance problems can appear caused by the wrong planning process of the bicycle
and vehicle roads. In developing and developed countries, strategic urban transportation planning is first goal to improve pedestrian
and bicycle transportation and reduce automobile utilization (Iacono et al., 2010; Litman et al., 2000). However, in these countries
where there is a lot of car dependence, automobiles not only are the focus of transportation systems, but also they generally push the


Corresponding authors.
E-mail addresses: meltemsaplioglu@sdu.edu.trv (M. Saplıoğlu), metinmutluaydin@gmail.com (M.M. Aydın).

https://doi.org/10.1016/j.jth.2018.05.011

Available online 21 May 2018


2214-1405/ © 2018 Elsevier Ltd. All rights reserved.
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

planning decision making processes (Newman and Kenworth, 1999; Balsas, 2003). In other words, motorized priority transport
planning is a misplaced policy. This wrong planning policy causes low bicycle access and makes less bike ownership (Pucher and
Buehler, 2008; Feleke et al., 2017; Lindsay et al., 2009).
From a different perspective, as a positive development, the budget allocated to transport infrastructure is increasing in devel-
oping countries day by day. For example, as a developing country, total transport infrastructure investment in Turkey has shown 21%
increase yearly from 2005 to 2015 compared to previous years (OECD, 2017). This significant budget increase supports the con-
struction of more safer and convenient routes. On the other hand, unlike large-scale cities, priority in small-sized cities is usually
given to vehicle route planning that shows the bicycle route safety is pushed to the second plan. This deficient planning policy is also
a common problem in countries like UK and USA because of the giving green light to private car according to urban transport policies
(Pucher and Buehler, 2008). Whereas, a good planning requires an integrated package of many different, complementary inter-
ventions, infrastructure provision and pro-bicycle programs, bicycle supportive land use planning and restrictions on car use (Pucher
et al., 2010). Such proper urban transportation planning policies are also encountered in the Netherlands, Germany and Denmark
(Pucher and Buehler, 2008). Studies showed that cities with high levels of bicycling and good safety rates tend to have extensive
infrastructure as well as pro-bicycle policies and programs. Whereas those with low bicycling rates and poor safety conditions
generally have done much less (Pucher and Dijkstra, 2003; Pucher and Buehler, 2008). For this reason, in-depth research for bicycle
route planning and the construction of bike routes is essential. In order to assign safe and convenient routes and improve existing
routes in a short period of time, Geographic Information System (GIS) can be used actively. Besides, the database to be obtained with
GIS will form an updatable interface for querying all the parameters required in the route selection.
This paper is guided by one major research question that how can safe and convenient bicycle routes integrated with existing
public transport systems. The main purpose of the study is to find safer and serviceable bicycle routes for bicyclists by using a safe
route selection exercise in which bicycle-bus integrated system route choice is based on the preference of cyclists and effective
parameters in the literature. To achieve this aim and determine important factors for choosing safer and serviceable routes, firstly a
questionnaire survey (QS) was carried on 460 participants who are cycling for a long time in the city center of Isparta/Turkey.
Secondly, a multi-criteria decision making approach Analytic Hierarchy Process (AHP) method is implemented to examine the effects
of parameters obtained from the survey data. Thirdly, accident prone locations have been determined from fatal and injury traffic
accident reports of the years 2016 and 2017 taken from Traffic Police Division of Isparta (TPDoI) and analysed by using Geographic
Information System (GIS) program of ArcGIS® 10. Then existing bicycle paths and bus stops of Isparta have been processed on a
digital map and safely integrated bicycle-bus routes are tried to obtain by taking results derived from the surveys and accident-prone
sections into the consideration. Finally, it is suggested to reduce bicycle driving stress in vehicular traffic, by reducing the distances in
dangerous parts of bicycle trips.

2. Literature review

2.1. An overview of bicycle-public transport integration types

Nowadays, cycling is used in urban transportation systems and it always interacts with other transportation modes. Bicycle-public
transport integration is an essential alternative for bicyclists to safe utilization at intersections with other transportation modes.
Because, it is observed that bicycle transport can be made safer and more comfortable when the safe utilization of bicycle coordinated
with public transport is examined (Pucher and Buehler, 2008). In practice, the integration of cycling with public transport systems
appears in many different ways. For example, bicycle-on-ferry services established in 1940s and 1950s were the first formalized co-
used programs for cycling and public transport. The process was followed by bicycle-on-bus services in 1970s and bicycle-on-rail
services in 1980s (Schneider, 2005). But recent evaluations on bicycle integration systems are researched as the bicycle on public
transport systems or bicycle parking programs coordinated with public transport systems. Studies of bicycle-on-bus services have
been done in communities such as Bremen/Germany (Brunsing, 1997). Also, there are some studies about bicycle-on-rail services in
Stuttgart/Germany (AASHTO, 1997) and Santa Clara Valley/California (Jenkins, 2001). Most of these conducted studies generally
have addressed the safety of bicyclists on light rapid transit (McClintock and Morris, 2003; Schneider, 2005). These studies aimed to
increase the utilization of public transport and bicycle utilization, but few of the studies have taken a comprehensive view and
detailed analysis in urban spatial area (Schneider, 2005). Then to fill this literature gap, Krizek and Stonebraker (2011) described and
assessed common bicycle and transit integration strategies. Study findings showed that the cost-effectiveness measure suggests that
enhancing bicycle parking at a transit stop.
Public Bike Sharing Systems (PBSS)is a common application to increase the bicycle utilization in many countries (Shaheen et al.,
2013). These systems provide editing the missing link between existing points of the public transportation and desired destinations by
offering a new form of mobility (Midgley, 2009). Similarly, there are also some remarkable studies about the performance measure of
integration between bike sharing and public transport systems (Keijer and Rietveld, 2000; Pucher and Ralph, 2009). For example in a
study, Keijer and Rietveld (2000) examined the travel time effectiveness of a bicycle when it is used coordinated with public
transportation. According to study results, it was claimed that the utilization of a bicycle both in access trips (from the home to the
station) and egress trips (from the station to the end of a trip) causes a significant reduction of door-to-door travel time of the trips
with public transport. Moreover, integration of bike and public transport could help cyclists under the conditions of “bad weather,
difficult topography, gaps in the bike-way network, and mechanical failures” (Pucher and Ralph, 2009). In another study, Shaheen
et al. (2013) emphasized that best station settlement, safety and technological management issues are important parameters between
bike sharing systems and public transport integration. The relationship among behaviors of bicyclists, route preferences, and its

237
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

determinants have great importance on the bicycle-public transport integration. In other words, it was seen from the previous studies
that there is a strong relation between bicyclists' route preference and its determinants (Kassoff and Deutschman, 1969; Koppelman
and Bhat, 2006; Sener et al., 2009). The importance of bicyclists’ ideas are emphasized in these studies. Public transport system
administers and city planners have considered cycling as feeder and distributor service which play a key role in terms of trans-
portation and it has been accepted as a common tool for analysis according to these survey analysis. For example, in their study Sener
et al. (2009) used data in their analysis from a survey of Texas/USA bicyclists. The results showed that a strong importance for both
route-related attributes and bicyclists’ demographics in bicycle route choice decisions.

2.2. Effective factors for selecting safe bicycle routes

The safe and convenient selection of bicycle routes in bicycle and public transport integrated systems will also increase the safety
of the system. In this study, effective factors for the bicycle route selection have been focused on and examined in detail by using the
literature. It was seen that there are many effective factors for selecting a safer and serviceable route. For e.g., in Asgarzadeh et al.
(2016)'s research, bicyclist injury severity is studied in bicycle-motor vehicle crashes at different intersections with different motor
vehicles. It was observed that when a bus was involved in a bicycle-motor vehicle crashes, the injuries of bicyclists were twice as
likely to be severe as compared with cycle-motor vehicle crashes involvement of smaller cars, vans or pickup trucks. In another study,
Baumann et al. (2012) performed a survey study with bus drivers and bicyclists about bus–bicycle interactions. It was obtained that
the overtaking of a bicycle by a bus was considered as an uncomfortable maneuver for both parties. Also, it was found that 68% of bus
drivers and 59% of bicyclists felt uncomfortable while interacting with each other. This finding is confirmed by a stated preference
field survey on bicycle infrastructure preferences (De Ceunynck et al., 2017). Addition to these findings, these researches imply that
close-passing of buses can create a subjective experience of being unsafe with bicycles (Guthrie et al., 2001; Parkin et al., 2007).
Bicycle parking facility is one of the important and effective factors on bicycle utilization. Bicycle-on-bus services mostly allow
bicyclists to mount their bicycles on front of the bus in many countries like USA (Schneider, 2005). Most of those buses carry two
bicycles by the help of bicycle racks. Although there are some bus services that allow bicyclists to get on the bus with their bicycles, it
can often be restricted not to make inside more crowded bicycle. For this reason, bicycle parking is an important problem for crowded
locations (Schneider, 2005). In Taylor and Mahmassasni (1996) survey study, they used hypothetical scenarios in a stated-preference
survey and it was asked respondents to rank their priorities for making a trip by bike. It was understood from the study that bicycle
lockers are the preferred parking facility to increase the bike utilization. Also, Uz (2003) reported that there should be built safe bike
parks in order to prevent bike was stolen which affects bicyclists route selection. As a suggestion, it was stated that it is necessary to
build a secure parking place against the bicycle grievance and to install locked bicycle lockers in the determined points of the city
(Uz, 2003).
Infrastructure of bicycle road also needs appropriate grade, suitable and separated lane and continuity of bike path or route. Since
bike utilization is entirely dependent on the physical condition of the driver, cycling on steep inclines is quite tiring. However, due to
the technological developments in the bicycle sector, effect of topography conditions is decreasing day by day. On the other hand,
while the infrastructure is being built, the longitudinal slope should not exceed certain values in terms of road geometry. For
example, Uz and Karasahin (2004) pointed out that road grade has an importance for bicyclists and suggested optimal average
maximum longitudinally grade is 5%. However, in some cases longitudinal grade and the distance range that can be applied when the
longitudinal inclination is applied more than 5% (Uz and Karasahin, 2004). For safe and convenient bicycle utilization, maximum
distances over 5% longitudinal grade are given in Table 1.
While selecting routes, bike lanes are considered as one of the most important criteria for both cyclists and motor vehicle users.
They encourage obedience with traffic laws and reduce conflicts (Hunter et al., 1999) as well as potentially reduce accidents (Lott and
Lott, 1976; Moritz, 1997; Teschke et al., 2012). According to DiGioia et al. (2017)’s study, maintaining a physical separation between
bicycles and motor vehicles (space and/or time) prevent the two modes from colliding; likewise separated bikeways which increase
the reaction time for both modes to avoid an impending collision. Consequently, bicycle and vehicle visibility will increase and give
drivers and bicyclists more reaction time. So separated bike lanes or cycle tracks provide safer riding facilities. Also, it is important to
emphasize that separated bike lane studies showed an effective reduction in crash rates (Lusk et al., 2013; Teschke et al., 2012).
Several urban streets with or without separated bike lanes are studied by Kroll and Ramey (1977). The results showed that when
bicycle lanes are present, the driver's overtaking behavior and fewer close overtakes are observed in transitions and fading behaviors.
In the study conducted by Stewart and McHale (2014), it was found that areas with no bicycle lanes on urban roads significantly

Table 1
Maximum distances for safe and convenient bicycle utilization over 5%
longitudinal grade (AASHTO, 1999; Uz, 2003).
Longitudinal slope (%) Maximum distance (m)

5;6 240
7 120
8 90
9 60
10 30
11 and more 15

238
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Table 2
Effective factors while selecting bike route obtained from literature.
Effecting factors Sources of Effective Parameters

Bicycle Parks Taylor and Mahmassasni (1996); Uz, (2003); Schneider, (2005)
Accident Prone Areas Noland and Kunreuther, (1995); Loo and Tsui, (2010); Yang et al., (2010); Ghasemlou et al., (2015); Chen, (2015)
Bus Line Asgarzadeh et al., (2016); Stewart and McHale, (2014); Baumann et al., (2012); De Ceunynck et al., (2017); Guthrie et al., (2001);
Parkin et al., (2007); Caulfield et al., (2012)
Connected Bike Lane Titze et al., (2008); Cervero et al., (2009); Habib et al., (2014)
Road Side Car Park Stewart and McHale, (2014); DiGioia et al., (2017); Hunter et al., (1999); Johnson et al., (2013); Teschke et al., (2012)
Separated Bike Lane Hunter et al., (1999); Kroll and Ramey (1977); Aultman-Hall and Hall, (1998); Love et al., (2012); Stewart and McHale, (2014); Lott
and Lott, (1976); Moritz, (1997); Teschke et al., (2012); DiGioia et al., (2017); Lusk et al., (2013); Teschke et al., (2012)
Traffic Capacity Uz, (2003); Yuksel Project, (2001); Eryigit, (2012)
Road Grade AASHTO, (1999); Uz and Karasahin, (2004); Apasnore et al., (2017)
Signalized intersection NCC, (2006); Davies et al., (1997); Garder et al., (1994); Du et al., (2015); Schepers et al., (2011); Monsere et al., (2012); Jensen,
(2008); Harris et al., (2013)

increased over taking distance compared to areas with bicycle lanes. In another studies, Titze et al. (2008) and Cervero et al. (2009)
found a statistically significant relationship between bicycle use and bike lane connectivity in recent investigations (Habib et al.,
2014). Vehicle traffic capacity has a great importance when any precautions are not taken for bicyclists and bicycle roads are not
regulated forr motorcycles and motor vehicles to share the same road surface. A motor vehicle which focus on to overtake the cyclist
should do just the same as overtaking another vehicle while overtaking the next lane. In this kind of traffic cases (mix traffic case),
bicyclists must share the same roads with motor vehicles whose physical characteristics such as speed, weight and maneuverability
are very different from each other. Also they are the most effected users in a traffic accident because they are vulnerable in the event
of a collision. Therefore, bicycle utilization is risky in high motor vehicle capacity roads as well as utilization demands are low in
roads with high traffic volume and not separated bicycle lanes (Uz, 2003; Yuksel Project, 2001).
Another problem for bicycle utilization is the presence of road side car parking. According to Stewart and McHale (2014)’s results,
statistical tests showed that presence of road side car parking in urban area is more important than the presence of cycle lanes because
of its effects on distance for a driver when overtaking a cyclist. So the presence of road side parking can also be found significant and
logical in terms of its effect. Furtherly, underscore that road side car parking appears to be hazard due to the cars crossing the cyclists’
space to leave a parking place and the potential of having a car door open directly in a cyclist's path (DiGioia et al., 2017; Hunter
et al., 1999; Johnson et al., 2013; Teschke et al., 2012).
It can be seen from all the investigated literature that the criteria affecting the route choice vary according to the demographic
characteristics of the bicycle users and locations. But, in this study, it was tried to find the general effects of parameters related to
safety that can be effective in route selection by investigating literature. It was found that there are many factors that can be effective
in safe cycling route selection as given in Table 2. Also in the study, the effective parameters are considered in two groups as positive
effect and negative effect due to the literature interpretation results. For example, cyclists do not prefer routes with busy bus lines,
busy roadside park routes, heavy traffic volume roads, the high longitudinal slope roads and accident prone areas or these routes are
not safe for their cycling. For this reason, the effects of these parameters are reflected as the negative effect, while selecting route.
Additionally, connected bike lane, signalized intersections, bicycle parks or separated bike lanes are preferred for selecting safe and
convenient routes. So these parameters have a positive effect in the route selection study. In addition to these, for safe and suitable
bicycle route, admittedly assessing all factors' importance together is difficult due to the large number, so it is understood that this is a
multi-criteria decision problem.

3. Analytic Hierarchy Process (AHP) method

Understanding the influence of each factor on the significance of route selecting is necessary. For this reason, a multi-criteria
decision making methods have been examined commonly. A Multi-Criteria Decision Analysis (MCDA) can solve the problem while
multiple criteria affecting to it. There are many MCDA algorithms, but mostly they are divided into Multi-Attribute (MADA) or Multi-
Objective (MODA) decision techniques. The Analytic Hierarchy Process is known as a MADA technique (Malczewski, 2006) and AHP
is first developed and used by Saaty in 1970. AHP is one of the most effective multi-criteria evaluation methods. It is commonly used
in a wide range of practical applications in a variety of decision-making processes (Murat. et al., 2015). AHP is useful for examining
the relationships among the components of a problem to take the proper decision (Papaioannou et al., 2015; Boroushaki and
Malczewski, 2010; Chen et al., 2011). A standout amongst the most conspicuous properties of the Analytic Hierarchy Process model is
its capacity to assess quantitative as well as qualitative criteria and alternative options on the equal preference scale (Kumar et al.,
2017). To organize the criteria into a hierarchical structure, complex decision problems can be examined through a framework that
combines GIS and AHP. Decision problems that require spatial data and knowledge are called spatial decision problems (Rajabifard
et al., 2003).They develop the spatial decision systems and have been integrated with geographic information systems in most studies
(Goodchild, 1992; Silva et al., 2014).
Many surveys on many different scientific disciplines have been widely applied referring to the GIS based AHP applications during
the last decade (Papaioannou et al., 2015). Integrated AHP and GIS have been applied in the land use and transportation planning
process through route optimization, evaluation of transportation networks and infrastructure land use selection (Banai, 2006; Wang

239
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Table 3
AHP pairwise comparison matrix for criteria (Vargas, 1990; Col and Gercek, 2014).
Criteria 1 Criteria 2 … Criteria n

Criteria 1 W1/ W1 W1/ W2 … W1/ Wn


Criteria 2 W2/ W1 W2/ W2 … W2/ Wn
: : : : :
Criteria n Wi / W1 Wi / W1 … Wi / Wn

et al., 2009; Xu et al., 2012). GIS, when applied with AHP, supplies a valuable tool for organizing a problem and displaying the
alternatives visually by using a collaborative decision-making (Malczewski, 2006). As a result, there is a necessity to continue to
apply this integrative decision-making environment to real-world scenarios such as urban transportation planning and selecting safe
routes.
In AHP application, weighted arithmetic means of data can be calculated easily. For instance: If survey results have the given
scores [ X1 , X2 , …, Xn ] and the number of bicyclists are [ A1 , A2 , …, An ], weighted arithmetic mean (w) can be used as given in Eq. (1).
A1 X1 + A2 X2 +...+ An Xn
w1 =
A1 + A2 +...+ An (1)
In AHP calculations firstly, the unstructured problem is defined and the type of information required is concluded. Secondly, the
decision hierarchy is also developed from the top to the lowest level. In the third level, the elements are systematically evaluated by
comparing them to one another two at a time with respect to their impact on an element above them in the hierarchy. Synthesizing
the pair-wise comparison matrix is done as given in Table 3. Table 3 shows the pairwise comparison scale used in the AHP developed
by Saaty (1977) for converting the importance of effecting factors’ judgements into numerical values and for computing the priorities
of the elements.
a a … … … … a1n
⎡ 11 12 ⎤
⎢ a21 a22 … … … … a2n ⎥
⎢ ⋮ ⋮ ⋮ ⎥
A=⎢ ⋮ ⋮ ⋮ ⎥
⎢ ⋮ ⋮ ⋮ ⎥
⎢ ⎥
⎢ ⋮ ⋮ ⋮ ⎥
a a
⎣ n1 n2 … … … … ann ⎦

Where; A is the pairwise comparison matrix, aij shows the pairwise comparison rating between the element 1 and j of a level with
1
respect to the upper level. aij (entries) are governed by the following rules aij > 0 ; aij = a , aij = 1.
ji
The relative weight value of each criterion is estimated by calculating eigenvalues and eigenvectors (Kumar et al., 2017; Col and
Gercek, 2014). For checking the Consistency Ratio (CR) of the judgments, some operations should be performed. Calculated value of
CR by using Eq. (2) where CI is the consistency index and RI is the random consistency index, should be less than 0.1 (Saaty, 1977). If
the consistency rate is above 0.1, the evaluation will not show sufficient consistency.
CI
CR =
RI (2)
The deviation from consistency is measured by consistency index as given in Eq. (3) where λ is the largest eigenvalue and n is the
order of the comparison matrix (Saaty, 1977).
λ−n
CI =
n−1 (3)
Randomly generated matrices form of the random consistency index which depends upon the number of elements being compared
as given in Table 4 (Saaty, 2000; Malczewski, 1999; Ozturk and Batuk, 2010).

4. General properties of pilot city: Isparta

Most of the cities in Turkey have convenient conditions for bicycle utilization according to their climate and topography. For
these reasons, it was expected a high-level of cycling in Turkish cities. Unfortunately, the cycling level in cities are very far behind the
expectations because of the planning mistakes (Ozgurluk, 2016). Meanwhile, Turkish Republic Ministry of Public Works and Housing
(TR-MoPWH) currently has decided that urban transport area should be prepared in accordance with the idea of integration of urban

Table 4
Randomness indicator (Saaty, 2000).
n 1 2 3 4 5 6 7 8 9

RI 0.00 0.00 0.58 0.90 1.12 1.24 1.32 1.41 1.45

240
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Fig. 1. Location of selected Pilot city (Isparta) in Turkey.

transportation plans. Also, it was decided that planning principals based on the urban development models that support sustainable
urban development, encourage the use of public transport (Turkish Min. of Pub. Works and Sett. Coun. of Urb., 2009). One of the
issues to be considered in urban transportation is the integration of public transport with the other transportation modes. In this
study, Isparta City/TURKEY (one of the most suitable cities) was selected as pilot city to examine cycling and public transport system
integration because of its proper climate, topography and high car ownership rate (Fig. 1). In Isparta, only buses are used for the
public transportation and bus stops have no connection to the bike paths, pass through the high-traffic on main arterials. The total
bike paths’ length is about 9 km and they are not linked to each other. Only 2.2 km of them are separated as bicycle lane; 6.8 km of
them are bicycle lanes with marked in mixed traffic conditions. Bicycle lanes with marked cannot be used efficiently by bicyclists
because of the car parking and this situation causes some important problems about bicyclists’ safety.
All bicycle lanes and bicycle paths were built to Isparta in 2014. But most of them are not linked to each other and they do not
have sufficient length for transport or most of them are far away from bus stops. Also most of the bicycle lanes are not be able to use
because of the marked bicycle lanes on the roads caused by the parallel parking of motor vehicles.

5. Application of research methodology

The methodology of the study has divided into three simple steps to explain and show process steps of used method clearly as
given in Fig. 2 and explained as follows:

5.1. Step 1: Definition of problem and determination of criteria

To determine the problems, opinions and suggestions of bicycle users, a survey which can constitute basis for a new safe bicycle
route integrated with public transport plan in Isparta city, has been carried out. The survey particularly guided by one major research

Fig. 2. Flowchart of the used methodology.

241
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

question that how can be the safe and convenient bicycle paths coordinated with bus lines. In other words, this survey is necessary to
find out the factors which are important for safe bicycle routes and the inadequacy of present bikeways. On the north-south direction,
the city center and university campus are considered as origin and destination points for bicycle bus integration routes. For large
populations, it was suggested that the Eq. (4) to yield a representative sample for proportions (Israel, 1992). For this purpose Eq. (4)
is used while determining the number of questionnaires applied to users in the study area:

t 2*p*q
n=
D2 (4)

where; n is Sample size (number of survey), p is estimated proportion of an attribute (bicycle utilization proportion for this study)
that is present in the population, q is 1-p, D is desired level of precision and t is abscissa of the normal curve that cuts off an area α at
the tails (1-α equals the desired confidence level, e.g. 95%) (Israel, 1992).
According to conducted traffic counting studies on main arterials of city center, bicycle journeys have found as 3% proportion for
102th Street; 5% proportion for 101th Street; 2% for 124th Street (Saplioglu and Uysal, 2015). According to these different results, it
is understood that there is a large vehicle population but the variability for bicycling rate is not exactly known. Therefore, p = 0.5
(maximum variability) is used to avoid bias while extrapolating. The value of t is obtained in statistical tables which contain the area
under the normal curve (while significance level is 95% and ± 5% precision, t is 1.96). According to analysis results minimum survey
numbers were calculated as given below:

(1, 96)2*0, 5*0, 5


n= = 385 (Minimum 385 survey should be applied)
0, 052

In the study, bicycle users are selected from students and staff of Suleyman Demirel University having a population of 6500
students. Survey has been put into practice for 500 bicycle users from university students and staff. Missing surveys are eliminated
and 460 questionnaires are put into practice for analysis. In the survey process, it was preferred to survey by hand to hand. All
obtained descriptive statistics of Isparta Bicyclists’ Survey are summarized in Table 5. From the survey with respect to their ages,
participants are distributed as; 56.3% aged between 21–30, 22% aged between 31–40, 1.5% aged between 51–60, 17% aged between
11–20, and 3.3% aged between 41–50. Regarding the results, there are no participants aged 60 or over and aged 10 or below. The
participants are asked questions such as: while using bicycle which kind of bicycle route they use; if they think it would be useful to
integrate cycling with public transport or not. It was seen that 64% of the bicyclists use originally common vehicle roads as an
obligation. A safe bike path separated from motor vehicle traffic is utilized only by 12% of bicyclists. When the bicycle users are asked
about the integration of cycling to public transport, a majority of them gave a positive response in this regard.

5.2. Step 2: Parameters’ Weight Calculations with AHP

Effective parameters on bicycle-bus route integration was obtained from the literature review as given in Table 2. Survey
questions were prepared by taking consideration of all these effective parameters. In the study, it was desired to prioritize these
effective parameters by using the bicycle users' preferences with the questionnaire. Therefore some questions were asked to identify
priority ranking of effecting factors on bicycle route safety and public transport integration for bicyclists. While performing the
survey, each bicyclist was asked for choosing (ranking) one of the scales 1, 3, 5, 7, 9 (1: Equal, 3: Moderate Importance, 5: Strong

Table 5
Descriptive statistics of Isparta city bicyclists' survey.
Parameter Description Percentage (%)

Gender Female 16.7%


Male 83.3%

Educational Level Bachelor's Degree 62.8%


Below Bachelor's Degree 37.2%

Employment Not Employed 43.9%


Employed 56.1%

Used Routes Common Vehicle Roads 64.0%


Separated Lane with Marking 10.0%
Separated Lane with Separator 5.0%
Partly on the Side Walk 14.0%
Completely on the Sidewalk 7.0%

Necessity of Integration Yes 70.0%


No Idea 23.9%
No 6.1%

Avg. Age (N ) = 27.0

Avg. Income ($ ) = 829.5

Note: These statistics are based on the sample. They do not present the weighted values.

242
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Fig. 3. Global percentages of responses given for the sub-criteria and the suitable-safe bicycle routes.

Importance, 7: Very Strong Importance, 9: Most Strong Importance). Obtained ranking results were given in Fig. 3.
As summarized in the last part of Section 2.2, negative and positive effects of all the criteria were found from the literature and
selected criteria from the literature (see in Table 2) were arranged in the direction of the responses from different cyclists' surveys.
However, recognizing and evaluating of the survey results and criteria rankings shown in Fig. 3 are extremely complex and difficult to
regulate these complexities for the optimal decision. For this purpose, it needs emerging of the results with decision-making tech-
niques. For the determination of the optimal decision, the importance of elements for the decision problem should be determined as a
priority. For example, AHP is chosen in this study because the prominent part of the AHP over other multi-criteria decision methods
take into account the decision maker's intuitive knowledge into the analytical decision (Saaty, 2000). In the decision process, bi-
cyclists were used as experts with their survey results. The importance value scale of 1, 3, 5, 7 and 9 (explained before) is necessary
for construct a pairwise comparison matrix of a set of criteria and Saaty (2000)’s importance or preference value scale has been used
in AHP model. Fig. 4 shows the used AHP model and its parameters.
In model, Longitudinal Slope of Road (Road grade) (RG), Separated Bicycle Lane (SBL), Connected Bicycle Lane (CBL),Traffic

Fig. 4. Used Analytic Hierarchy Process (AHP) Model.

243
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Table 6
AHP pairwise comparison matrix using survey results.
— APA BL RSCP BP RG S TC CBL SBL

APA 1 1.04 1.06 1.13 1.27 1.35 1.45 1.58 1.83


BL 1 1.02 1.08 1.22 1.30 1.40 1.52 1.76
RSCP 1 1.06 1.19 1.27 1.37 1.49 1.72
BP 1 1.13 1.20 1.29 1.41 1.62
RG 1 1.07 1.15 1.25 1.44
S 1 1.07 1.17 1.35
TC 1 1.09 1.26
CBL 1 1.15
SBL 1

Capacity (TC), Bicycle Parks (BP), Bus Line (BL), Accident Prone Areas (APA), Road Side Car Park (RSCP) and Signalization (S) are
considered effective factors. In addition to this, the sum of the total weights of all criteria is determined as equal to one as seen in Eq.
(5),
WRG + WSBL + WCBL +WTC + WBP + WBL + WAPA + WRSPC + WS = 1 (5)

Binary comparison ratios of weight values before normalization are given in Table 6.
The value range for the upper triangle is between about 1 and 1.83. The weighting must be between 1 and 9. For this reason, value
1 was assigned to 1 and value 9 was assigned to 1.83. Other numbers have been assigned in proportion to the values in between
(Table 7). The lower triangular values of the matrix are the upper triangular values of the matrix divided by 1.
Then the local weights in Table 7 are found by summing the rows after just described the weighted matrix is normalized. In other
words, the super matrix weighted by the eigenvectors that are obtained by MatLab® software and results are given as weight in
Table 8.
Also, the sensitivity analysis is performed for AHP model (Saaty. 2000). A consistency ratio was recommended by Saaty (2000)
has been used to measure the consistency of the comparison. The weight values and the columns of the binary comparison matrix
were multiplied in order to indicate the consistency ratio. The sum of row in the matrix was obtained according to the obtained
values. This (nx1) weighted total vector was divided by the weight values to obtain the consistency vector. The consistency index (Eq.
3) is calculated based on the average value of the consistency vector and the number of criteria. The consistency rate (Eq. (2)) is
obtained by dividing the consistency index with the randomness indicator (Table 4) which varies depending on the number of
comparison criteria.

5.3. Step 3: Locating all effecting parameters into the ArcMap, making queries and finding safe bicycle roads

GIS is usually used for geographic data acquisition and processing. Many decisions such as site selection, planning evaluation of
transportation networks and route selection are concerned with spatial location as mentioned in Section 3. In this study, problem has
been modelled in hierarchically and the weights of the criteria forming are calculated with the AHP by using spatial decision analysis.
While applying the spatial decision analysis, weighting and ranking methods are found more suitable for a large number of criteria
(Malczewski. 1999; Articte. 1995). There were many numbers of criteria in the study. For this reason, weighting and ranking methods
are used in attribute table queries of ArcMap® 10 to evaluate them.
Isparta province roadmap was used in the study. This map is prepared in vector data structure and ArcMap® 10. The parameters
used for bicycle path selection were arranged in such a way that the map used in the attribute table of the road layer. In other words,
the effective parameters were processed to the attribute table of the road layer. Subsequently, the queries based on the attribute data
prepared by using Structural Query Language called SQL. It is the ideal tool to access and manipulate data in SQL databases (Kupcu
et al., 2015). One of the most powerful features of GIS is the ability to access map elements from attribute data or to access attribute
data from map elements by using queries. Query based on attribute data are standard database queries. Selected records from

Table 7
AHP pairwise comparison matrix of criteria scaled to 1–9.
– APA BL RSCP BP RG S TC CBL SBL

APA 1 2 2 3 3 4 5 7 9
BL 0.50 1 2 2 3 4 5 6 8
RSCP 0.50 0.50 1 2 3 4 5 6 8
BP 0.33 0.50 0.50 1 2 3 4 5 7
RG 0.33 0.33 0.33 0.50 1 2 2 3 5
S 0.25 0.25 0.25 0.33 0.50 1 2 2 4
TC 0.20 0.20 0.20 0.25 0.50 0.50 1 2 3
CBL 0.14 0.17 0.17 0.20 0.33 0.50 0.50 1 2
SBL 0.11 0.13 0.13 0.14 0.20 0.25 0.33 0.50 1

244
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Table 8
Criteria weights and result data.
– APA BL RSCP BP RG S TC CBL SBL Weight

APA 0.297 0.394 0.304 0.318 0.222 0.208 0.201 0.215 0.191 0.261
BL 0.148 0.197 0.304 0.212 0.222 0.208 0.201 0.185 0.170 0.205
RSCP 0.148 0.099 0.152 0.212 0.222 0.208 0.201 0.185 0.170 0.177
BP 0.099 0.099 0.076 0.106 0.148 0.156 0.161 0.154 0.149 0.127
RG 0.099 0.066 0.051 0.053 0.074 0.104 0.081 0.092 0.106 0.081
S 0.074 0.049 0.038 0.035 0.037 0.052 0.081 0.062 0.085 0.057
TC 0.059 0.039 0.030 0.027 0.037 0.026 0.040 0.062 0.064 0.043
CBL 0.042 0.033 0.025 0.021 0.025 0.026 0.020 0.031 0.043 0.030
SBL 0.033 0.025 0.019 0.015 0.015 0.013 0.013 0.015 0.021 0.019

attribute table are also determined on the map by database inquiry. Thus the inquiry can be displayed on the map. In the study, the
queries based on the attribute data were made by questioning the literary data of the road layer in the map. For this aim, the slopes of
the streets are calculated step-by-step operations and queries are obtained as follows:
Firstly, a field added as “Road Grade” (RG) in the attribute table. Then DEM data-30 m were used (ASTER GDEM, 2017) in the
topographic information of terrain. Surface information was added using the 3D analysis surface function (3D Analysis Tools-
Functional Surface) in Arc Toolbox. The purpose in here is to transfer the slope of the terrain to the road data in a short time (Fig. 5).
In the attribute table, the slope values are visible under the “Average Slope” field (Fig. 5). Then, all street lengths on the map were
calculated by using Calculate Geometry in ArcMap Toolbox and added as a new field “Slope Length”. In the next step, the longitudinal
slope constraint information allowed for cyclists from the literature (see Table 1) was utilized. To get these constraints into account, a
new field “Road Grade” added. In accordance with the criteria given in Table 9, 1 or 0 values were written in this “Road Grade” field
in attribute table. More specifically, we should have chosen the routes that are appropriate to the bike's utilization slope. Thus in
“Average Slope” field some calculations have been made by using Field Calculator tool. Eq. (6) is one of the calculations made by
considering Table 1 and Field Calculator tool. Then. “1” is written to the selected features from the results of calculations.

(“Average Slope” ≥ 5 and“Average Slope” ≤ 6) and “Slope Length” < 240 (6)

Fig. 5. Isparta city slope and road map.

245
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Table 9
Weights of effective parameters for bicycle route selection.
Parameters Global weights

Accident Prone Areas (APA) 0.261


Bus Lane (BL) 0.205
Road Side Car Park (RSCP) 0.177
Bicycle Parks (BP) 0.127
Road Grade (RG) 0.081
Signalization (S) 0.057
Traffic Capacity (TC) 0.043
Connected Bike Lane (CBL) 0.030
Separated Bike Lane (SBL) 0.019

The same queries were made for the other values of Table 1 and the value “1” was written. The remaining lines are also filled with
“0”. Our goal is to write 1 and 0. If we write 1 these roads will get weights when we multiply the weights (in Table 8) which are
gathered from AHP results; but if we write 0, these streets will not get any weights about this field (Road Grade). Ranks of road grades
in Table 1 were taken into consideration and written with using field calculator queries. So all streets got a different average grade on
the map. Also, if the paths are within the permissible road grade and length values given in Table 1, then the incremental road grade
parameter will participate in the last analysis (see Eq. 8). For this reason, they are written as “1” in Road Grade field in the attribute
table.
Secondly, Accident Prone Area “APA” field is formed. Before forming the field APA, it is worth to explain that Isparta city has busy
motorized traffic and bicycle. Until recently, in the characteristics of bicycle crashes have received little attention by road safety
researchers because of the difficulty of getting historical data which are on the paper records and not linked with digital maps. For
this reason, gathering the years of 2016–2017 accident records were necessary on the digitized map. After getting the accident
records from Directory of Isparta Traffic Region, they were arranged and linked to Isparta digital map with ArcMap® 10. By producing
accident database and using rank query the number of fatal and injured accidents which occurred along road segments and inter-
sections over a two-year period were found. Fig. 6 shows analysis result of the distribution of accident locations in red coloured paths
and a bicycle accident photo was taken just after an accident at 102th street. 1221 fatal and injured accidents were observed in two
years’ time. 8 of them are fatal and 2 of the fatal bicycle accident occurred in 2016; 5 of them are fatal and none of the fatal accidents
is bicycle accident occurred in 2017. Not only intersection accidents but also accident-prone streets were found on the map by using
queries. Intersection accidents can be seen as point data. But for analysis, intersection accident numbers were added to roads which
the junction is connected because of all the queries made in the same layer. Then for the APA field, other steps are followed.

Fig. 6. 2016–2017 Accident locations and accident prone streets in Isparta.

246
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Fig. 7. a) Current bus links, b) bicycle lane and paths in Isparta city.

In the APA, effective parameters of the attribute table, calculated values between 0 and 1 values (for not the only 0–1 but also
interval values between 0 and 1). Because there were many values in accident-prone area criteria. In other words, the number of
accidents was normalized into specific number ranges. When the criteria are in different numerical ranges of measure, they must be
normalized to a standard number range. Eq. (7) is the most widely used method for this purpose referred as linear scale transfor-
mation (Malczewski. 1999; Articte. 1995). Linear scale transformation was used in this part to determine the maximum-minimum
value range of accident numbers. Then the values were written in the APA field of attribute table.

x ij − x j min
x ij =
x j max − x j min (7)

Thirdly, traffic capacity values were gathered from Isparta Municipality Traffic Department and processed to the street data table.
Traffic capacity criteria, such as accident prone location criteria, have numerical values. For this reason, they were normalized to a
standard number range and a linear scale transformation is used. Then the values were written in the Traffic Capacity field. Also, the
current locations of bus links and bicycle routes (lanes) were gathered from Isparta Public Buses Cooperative Establishment and they
were digitized as given on the map in Fig. 7(a-b), respectively.
After Road Grade, Traffic Capacity and Accident Prone Area were normalized and assigned to the fields, the other bicycle route
selection parameters (BL, RSCP, BP, S, CBL and SBL) were interpreted and recorded as values in the range of 0–1. For example, it has
been processed to the digitized map street database fields as: if there are separated bike lane on the street 1, otherwise 0; if there is
roadside car park on the street 1, otherwise 0; if there is connected bike lane on the street 1, otherwise 0; if there are bike parks on the
street 1, otherwise 0; if there is signalized intersection on the street 1, otherwise 0; if there is bus line on the street otherwise 1,
otherwise 0.

6. Research findings and discussion

The weight of bicycle path selection criteria ranked by using AHP as given in Table 9. It can be emphasized in Section 5 that
Accident Prone Area factor is found as the most important factor according to the weights of the criteria (APA) which has the highest
weight. This results support the Chen (2015)’s study that cyclists tend to stay away from dangerous locations. Because APA criteria is
an important criterion (Chen, 2015). Furthermore, it has been understood that Bus Lane (BL) and roadside car parking (RSCP) are
second and third essential criteria while selecting routes, respectively.
The criteria of Road Grade (RG) rank among fifth place in our study. It is known that the longitudinal slope of the literature was
influential for passing distance on cycling routes (Apasnore et al., 2017). But, it is surely that bicycle users in different regions and
cities will respond differently to some of the criteria. Because the ideas of cyclists will change according to place or terrain changes.
For example, if a terrain used by a bicycle is all flat area like in Isparta City, the longitudinal slope may not be attached much
importance to in the flat area while selecting bicycle-bus integrated route. It is important to add that when the slopes of the roads in
the province of Isparta were questioned in GIS, it was found that 3072 streets of 3953 had a small inclination from 4% of the
longitudinal slope. Likely for this reason, cyclists have not chosen longitudinal slopes criteria (RG) at first sight while choosing routes.
Also safe and suitable route selection in Isparta city should systematically be take place with respect to these rankings by using GIS.
But it should be kept in mind that the results are completely based on the questionnaires which were taken from bicycles located in
the city center of Isparta. Notwithstanding these limitations, the importance of Road Grade worth remains in a number of situations

247
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Fig. 8. Suggested safe bicycle routes.

but ranking of importance can be change. If the areas where the terrain is not flat, bicyclists care much more about the longitudinal
slope.
According to literature review, it has been determined that separated bicycle lane has a positive impact. Because separated bike
lane is safer than mix traffic conditions (DiGioia et al., 2017; Lusk et al., 2013; Teschke et al., 2012). On the other hand, bus lines
have a negative impact on the safety of bicycle movements. Because both bicycles and buses tend to get away from each other in
mixed traffic as can be seen from the literature (De Ceunynck et al., 2017; Caulfield et al., 2012; Guthrie et al., 2001; Parkin et al.,
2007). Road side car park has negative impact for bicyclists because of car drivers and passengers who carelessly open the door of the
parked vehicle (DiGioia, 2017; Hunter et al., 1999; Johnson et al., 2013; Teschke et al., 2012). While selecting bicycle routes, they
prefer routes with parking restrictions on the roadside. The other parameters (accident prone location, road grades and high traffic
capacity) have negative impacts on bicyclists. So their effects will be negative while making analysis. Despite that, connected bike
lane, bike parks and signalized intersections have positive effects. These negative and positive effects are considered while Eq. (8) is
calculated. In summary, Eq. (8) is used for reaching the Final Map (Fig. 8).
FINALMAP = [−(BL) x 0.205] + [−(RSCP ) x 0.177] + [(−(RG ) x 0.081)]
+ [−(APA) x 0.261] + [(BP ) x 0.127] + [(CBL) x 0.03]
+ [(S ) x 0.057] + [−(TC ) x 0.043] + [(SBL) × 0.019] (8)

Weights of criteria are multiplied by different criteria values in the fields in ArcMap that are calculated as roads of field data. In
the final map (Fig. 8), the values obtained from the multiplication results are divided into four groups as shown in the map legend.
While dividing the result values into four group, geometric interval of classification tool has been used in ArcMap. This classification
method was used for visualizing continuous data and it provides an alternative to the quintiles’ classification method. This is a
classification scheme where the class breaks are based on class intervals that have a geometrical series. The geometric coefficient in
this classifier can change according to class ranges. The used algorithm creates these geometrical intervals by minimizing the square
sum of element per class. This situation ensures that each class range has approximately the same number of values with each class

248
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Fig. 9. Current bus lines and APA and final map of suitable or unsuitable bicycle routes.

and that the change between intervals is fairly consistent. It produces a result that is visually appealing and cartographically com-
prehensive. It also minimizes variance within classes, and can even work reasonably well on examined data that is not normally
distributed (ESRI, 2009). This made it easier to comment on the obtained map. For example, range from −0.518 to −0.017 are
negative and coloured in red. It can be said that bicyclists do not prefer these roads. The range from 0.093 to 0.451 are selected routes
by bicyclists and they are coloured in green. When the bus line map and accident-prone area map and next to the final map examined
in Fig. 9. It is clear to see that the red line of the final map is sometimes the same lines with the bus lines and sometimes the same lines
with the APA lines. But, the other parameters in bicyclists’ opinion were used in the analysis although they had less effect. This
supports the idea that bicycle users in Isparta have not chosen for bus lines and accident-prone areas due to poor safety. As a result,
the roads with the least chosen are the ones with the negative result (red color); the roads with the highest degree of preference are
with the positive high-value results, considered as green roads.
At the same time, priority should be given to green roads in Isparta where bicycle routes will be planned in urban transportation

Fig. 10. Suggested bicycle-bus integration network.

249
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

planning which will support the bicycle integrated bus system. It is advisable to build separated bicycle paths where bus lines and
bicycle lines are coinciding. To find these paths, the layers of bus line and bike lanes are shown at the same time in Fig. 10. Blue roads
are the current bus lines. This map is similar to the finding of the bicycle routes approach in the literature which should be safe and
convenient. Because the green bicycle roads are mostly not the same route as accident-prone roads. Current bus lines and other roads
which have negative impact parameters. So dark green roads can be suggested as the most preferred bicycle roads and can be built
first.

7. Conclusion and recommendations

The importance of integrating cycling with public transport for the encouragement of cycling is being emphasized by various
countries in the world and in Turkey. But they are not sufficient in most of cities in Turkey in order to positively effect on this
situation and to establish a safer and suitable bicycle-bus integration routes. According to bicyclists, it is usually difficult to find
which route can be chosen for safe and suitable bicycling and which bus lines can be used for integration because of the lack of the
database. For this purpose, in this study a field survey has been carried out for exploring this insufficiency and its proportionality with
bicycle users’ need in Isparta city.
According to the survey and Analytic Hierarchy Process (AHP) results, Accident Prone Areas (APA) is the most important factor
for bicyclists. The other parameters are determined as Bus Lane (BL), Road Side Car Park (RSCP), Bicycle Parks (BP), Road Grade
(RG), Signalization (S), Traffic Capacity (TC), Connected Bike Lane (CBL) and Separated Bike Lane (SBL) according to their im-
portance, respectively. After weighted with AHP, Spatial Decision has been used with Geographic Information System (GIS) to
calculate data of all effective criteria and visualize the map. According to the results, new bicycle routes which are linked to bus lines
are formed. Although, it was selected nine parameters from the literature, these can be supported with different parameters according
to the change of the zone and the demand change.
Bicycle routes should be linked to this bus link. After the queries, new bicycle paths and integrated bus links are achieved as given
in Fig. 10. Green lines show safer and suitable bicycle paths whose slope are not higher than 5% in long distances and there can be
built integrated bicycle paths. Bicycle lanes and bus links can be connected at the bus stops where there are two huge bicycle parking
areas as seen in Fig. 10. It was found that in Isparta city only 102th street has a long continuing length and 5% slope and more than
240 m; the other street lengths are either short or the gradients are not more than 5%. Also not only the most bicycle accidents are
occurred in 102th street but also it is situated as a major road for bus links as in Fig. 9. So it is logical that the results give this street as
a bus line and not enough safe for bicyclists. Currently, there are only 2.2 km long separated bicycle lane in Isparta city center to
increase the length and to find the new safe separated bicycle lane locations. Fig. 9 can be used by municipalities. For example they
can use this map while deciding which route should be separated bike lane; which of these streets should be use first.
On the other hand study has some limitations. One limitation is the lack of existing research on cyclist behavior of different
cyclists whom use bike in any other cities any other countries. Due to these cultural differences, it is difficult to directly compare the
findings with that of similar studies carried out in foreign cities. In other words, cyclists answered only bicycle utilization in the city
center of Isparta. This may be a problem for further work in different cities. For example, it has been found that bicycle users are
significantly influenced by the Road Grade in the literature review. However, there are not many longitudinal slopes in the city center
of Isparta moderately important while selecting a bike route the selection of routes in the questionnaires. However, it is exactly that
bicycles can be influenced significantly from the longitudinal slope in cities with different topographical structures. This changes the
weight order. Also, weights may be specific to different regions based on the human nature and driver behaviors and even to the time
of day (in other words. different driving cultures. congestions or peak hour behavior). For this reason, it is proper to carry out
separate questionnaire surveys for that region in cycling route planning studies in different cities and regions to be made in the future.
In future studies, bicycle utilization can be integrated with all (bus. rail and ferry) the public transport modes in the same time. Thus,
more public transport or private car passengers can be encouraged about the utilization of the bicycle as a transport mode.

Acknowledgements

The study was funded by the Suleyman Demirel University Scientific Research Project and connected to cooperation with COST
Action Project TU1101 (SDU-BAP Project Management No: 3414-YL1-13): Towards Safer Bicycling through Optimization of Bicycle
Helmets and Utilization. We would like to thank COST Action Project TU1101-Work Groups; Isparta Police Department and Isparta
Municipal Traffic Office, Suleyman Demirel University Project Coordination Unit and Emel Yuzer Gunay for the kindly assistance.
The study is also deemed worthy of Transist Secure Service Project Award in Istanbul Transist 2015.

Conflict of interest

The authors report that they have no conflicts of interest.

References

AASHTO, 1997. Stuttgart LRT Gives Bikes a Push—Literally, Washington D.C.


AASHTO, 1999. Guide for the Development of Bicycle Facilities. American Association of State Highway and Transportation Officials, Washington, USA.
Apasnore, P., Ismail, K., Kassim, A., 2017. Bicycle-vehicle interactions at mid-sections of mixed traffic streets: examining passing distance and bicycle comfort

250
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

perception. Accid. Anal. Prev. 106, 141–148.


Articte, P.N., 1995. Raster procedures for multi-criteria/multi-0biective decisions. Photogramm. Eng. Remote Sens. 61 (5), 539–547.
Asgarzadeh, M., Verma, S., Mekary, R.A., Courtney, T.K., Christiani, D.C., 2016. The role of intersection and street design on severity of bicycle-motor vehicle crashes.
Inj. Prev. 2016–042045.
ASTER Global Digital Elevation Model (ASTER GDEM, 2017. Retrieved from: 〈http://www.jspacesystems.or.jp/ersdac/GDEM/E/4.html〉. (Accessed 15 August 2017).
Aultman-Hall, L., Hall, F.L., 1998. Ottawa-Carleton commuter cyclist on-and off-road incident rates. Accid. Anal. Prev. 30 (1), 29–43.
Balsas, C.J., 2003. Sustainable transportation planning on college campuses. Transp. Policy 10 (1), 35–49.
Banai, R., 2006. Transit station area land use/site assessment with multiple criteria: an integrated GIS-expert system prototype. J. Publ. Transp. 3 (1), 95–110.
Baumann, C., Brennan, T., Zeibots, M.E., 2012. Bike Rider and Bus Driver Interaction Study – Draft Report Prepared for the City of Sydney. Institute for Sustainable
Futures, University of Technology Sydney, Sydney, Australia.
Boroushaki, S., Malczewski, J., 2010. Using the fuzzy majority approach for GIS-based multicriteria group decision-making. Comput. Geosci. 36 (3), 302–312. http://
dx.doi.org/10.1016/j.cageo.2009.05.011.
Brunsing, J., 1997. Public Transport and Cycling: Experience of Modal Integration in Germany. Greening of Urban Transport (Chapter 27). John Wiley and Sons, Ltd,
Chichester, England, pp. 357–370.
Caulfield, B., Brick, E., McCarthy, O.T., 2012. Determining bicycle infrastructure preferences – a case study of Dublin. Transp. Res. Part D: Transp. Environ. 17 (5),
413–417. http://dx.doi.org/10.1016/j.trd.2012.04.001.
Cervero, R., Sarmiento, O.L., Jacoby, E., Gomez, L.F., Neiman, A., 2009. Influences of built environment on walking and cycling: lessons from Bogota. Int. J. Sustain.
Transp. 3, 203–226.
Chen, P., 2015. Built environment factors in explaining the automobile-involved bicycle crash frequencies: a spatial statistic approach. Saf. Sci. 79, 336–343.
Chen, Y.R., Yeh, C.H., Yu, B.F., 2011. Integrated application of the analytic hierarchy process and the geographic information system for flood risk assessment and
flood plain management in Taiwan. Nat. Hazards 59 (3), 1261–1276. http://dx.doi.org/10.1007/s11069-011-9831-7.
Col, Y.D., Gercek, H., 2014. Prioritization of integrated bicycle network clusters in istanbul using analytic hierarchy process. Pamukkale Univ. J. Eng. Sci. 20 (6),
215–224. http://dx.doi.org/10.5505/pajes.2014.29292.
Davies, D.G., Taylor, M.C., Ryley, T.J., Halliday, M., 1997. Cyclists at roundabouts-the effects of' continental' design on predicted safety and capacity. TRL REPORT
285.
De Ceunynck, T., Dorleman, B., Daniels, S., Laureshyn, A., Brijs, T., Hermans, E., Wets, G., 2017. Sharing is (s caring? Interactions between buses and bicyclists on bus
lanes shared with bicyclists. Transp. Res. Part F: Traffic Psychol. Behav. 46, 301–315.
DiGioia, J., Watkins, K.E., Xu, Y., Rodgers, M., Guensler, R., 2017. Safety impacts of bicycle infrastructure: a critical review. J. Saf. Res. 61, 105–119.
Du, Y., Wu, J., Qi, G., Jia, Y., 2015, March. Simulation study of bicycle multi-phase crossing at intersections. In: Proceedings of the Institution of Civil Engineers-
Transport Vol. 168, No. 5, pp. 457–465.
Eryigit, S., 2012. Sürdürülebilir Ulaşımın Sosyal Boyutunda Bisikletin Yeri (In Turkish, Phd Thesis). Selçuk University, Konya, Turkey.
ESRI, 2009. ArcGIS Desktop Help 9.32, 〈http://webhelp.esri.com/arcgisdesktop/9.3/index.cfm?TopicName=welcome〉 (Accessed 22 June 2009).
Feleke, R., Scholes, S., Wardlaw, M., Mindell, J.S., 2017. Comparative fatality risk for different travel modes by age, sex, and deprivation. J. Transp. Health. 1–14.
Garder, P., Leden, L., Thedeen, T., 1994. Safety implications of bicycle paths at signalized intersections. Accid. Anal. Prev. 26 (4), 429–439.
Ghasemlou, K., Aydin, M.M., Yıldırım, M.S., 2015. Prediction of pedal cyclists and pedestrian fatalities from total monthly accidents and registered private car
numbers. Arch. Transp. 34, 29–35.
Goodchild, M.F., 1992. Geographical information science. Int. J. Geogr. Inf. Syst. 6, 31–45. http://dx.doi.org/10.1080/02693799208901893.
Guthrie, N., Davies, D.G., Gardner, G., 2001. Cyclists' Assessments of Road and Traffic Conditions: the Development of a Cyclability Index No. TRL Report 420.
Transport Research Laboratory, Wokingham, UK.
Habib, K.N., Mann, J., Mahmoud, M., Weiss, A., 2014. Synopsis of bicycle demand in the City of Toronto: investigating the effects of perception, consciousness and
comfortability on the purpose of biking and bike ownership. Transp. Res. Part A: Policy Pract. 70, No.67–80.
Harris, M.A., Reynolds, C.C.O., Winters, M., Cripton, P.A., Shen, H., Chipman, M.L., Teschke, K., 2013. Comparing the effects of infrastructure on bicycling injury at
intersections and non-intersections using a case-crossover design. Inj. Prev. 19, 303–310.
Hunter, W.W., Stewart, J.R., Stutts, J.C., Huang, H., Pein, W.E., 1999.. A comparative analysis of bicycle lanes versus wide curb lanes: final report, tech. rep. FHWA-
RD-99-034. FHWA.
Iacono, M., Krizek, K.J., El-Geneidy, A., 2010. Measuring non-motorized accessibility: issues, alternatives, and execution. J. Transp. Geogr. 18 (1), 133–140.
Israel, G.D., 1992. Determining Sample Size. University of Florida Cooperative Extension Service., Institute of Food and Agriculture Sciences, EDIS.
Jenkins, A., 2001. Bikes on vta’s trains: a success story. Boston, Mass June 10-14, American Public Transportation Association, Rail Transit Conference, pp. 4.
Jensen, S.U., 2008. How to obtain a healthy journey to school. Transp. Res. part A: policy Pract. 423, 475–486.
Johnson, M., Newstead, S., Oxley, J., Charlton, J., 2013. Cyclists and open vehicle doors: crash characteristics and risk factors. Saf. Sci. 59, 135–140.
Kassoff, H., Deutschman, H.D., 1969. Trip generation: a critical appraisal. Highway Research 297. Highway Research Board, NW Washington. DC 20001, United States.
Keijer, M.J.N., Rietveld, P., 2000. How do people get to the railway station? The Dutch experience. Transp. Plan. Technol. 23 (3), 215–235.
Koppelman, F.S., Bhat, C.R., 2006. A self instructing course in mode choice modeling: multinomial and nested logit models. Prepared for U.S. Department of
Transportation. Federal Transit Administration. Washington, DC 20590 United States.
Krizek, K.J., Stonebraker, E.W., 2011. Assessing options to enhance bicycle and transit integration. Transp. Res. Record: J. Transp. Res. Board 2217, 162–167.
Kroll, B.J., Ramey, M.R., 1977. Effects of bike lanes on driver and bicyclist behavior. J. Transp. Eng. 103 (2), 243–256.
Kumar, S., Srivastava, P.K., Snehmani, 2017. GIS-based MCDA–AHP modelling for avalanche susceptibility mapping of Nubra valley region, Indian Himalaya.
Geocarto Int. 3211, 1254–1267.
Kupcu, S., Inceoglu, M., Uygucgil, H., Cabuk, A., Bektore, E., Cabuk, S., Comert, R., Isık, O., Ersoy, M., 2015. GeographicInformation Systems, ISBN 978-975-06-1791-
1, Anadolu University Book No: 3161, p 269, in Turkish.
Lindsay, R., G., MittalI, Mohan, D., Tiwari, G., Woodward, A., Roberts, I., 2009. Public health benefits of strategies to reduce greenhouse-gas emissions: urban land
transport. Lancet 374, 1930–1943.
Litman, T., Blair, R., Demopoulos, W., Eddy, N., Fritzel, A., Laidlaw, D., Maddox, H., Forster, K., 2000. Pedestrian and bicycle planning: A guide to best practices.
Victoria Transport Policy Institute.
Loo, B., Tsui, K.L., 2010. Bicycle crash casualties in a highly motorized city. Accid. Prev. 42 (6), 1902–1907. http://dx.doi.org/10.1016/j.aap.2010.05.011.
Lott, D.F., Lott, D.Y., 1976. Effect of bike lanes on ten classes of bicycle–automobile accidents in Davis, California. J. Saf. Res. 84, 171–179.
Love, D.C., Breaud, A., Burns, S., Margulies, J., Romano, M., Lawrence, R., 2012. Is the three-foot bicycle passing law working in Baltimore, Maryland? Accid. Anal.
Prev. 48, 451–456. http://dx.doi.org/10.1016/j.aap.2012.03.002.
Lusk, A.C., Morency, P., Miranda-Moreno, L.F., Willett, W.C., Dennerlein, J.T., 2013. Bicycle guidelines and crash rates on cycle tracks in the United States. Am. J.
Public Health 103, 1240–1248.
Malczewski, J., 2006. GIS‐based multicriteria decision analysis: a survey of the literature. Int. J. Geogr. Inf. Sci. 20 (7), 703–726. http://dx.doi.org/10.1080/
13658810600661508.
Malczewski, J., 1999. GIS and Multi-Criteria Decision Analysis. John Wiley Sons, Inc., New York.
McClintock, H., Morris, D., 2003. Integration of cycling and light rapid transit: realizing the potential. World Transp. Policy Pract. 9 (3), 9–14. (ISSN 1352-7614).
〈http://www.eco-logica.co.uk/worldtransport.html〉.
Midgley, P., 2009. The role of smart bike-sharing systems in urban mobility. Journeys 2 (1), 23–31.
Monsere, C., Figliozzi, M.A., Thompson, S., Paulsen, K., 2012. Operational guidance for bicycle-specific traffic signals in the United States.
Moritz, W.E., 1997. Survey of North American bicycle commuters: design and aggregate results. Transp. Res. Rec. 1578, 91–101.
Murat Y.S., Arslan T., Cakıcı Z., Akcam C., 2015. Analytical hierarchy process ahp based decision support system for urban intersections in transportation planning. In:

251
M. Saplıoğlu, M.M. Aydın Journal of Transport & Health 10 (2018) 236–252

Johnston, L. (ed.) Using Decision Support Systems for Transportation Planning Efficiency. s.l.:United States of America by Engineering Science Reference an
imprint of IGI Global. pp. 203–202.
NCC, 2006. Cycling Design Guide. Nottinghamshire County Council, West Bridgford, UK.
Newman, O., Kenworth, J., 1999. Sustainability and cities: overcoming automobile dependence.
Noland, R.B., Kunreuther, H., 1995. Short-run and long-run policies for increasing bicycle transportation for daily commuter trips. Transp. Policy 2 (1), 67–79.
OECD, 2017. 〈https://stats.oecd.org/Index.aspx?DataSetCode=ITF_INV-MTN_DATA#〉.
Ozgurluk, M.E., 2016. Developing an External Bike Rack Design for Inner City Public Busses through an Action Research Process. Diss. Middle East Technical
University.
Ozturk, D., Batuk, F., 2010. Analytic hierarchy process for spatial decision making. sigma. J. Eng. Nat. Sci. 28, 124–137.
Papaioannou, G., Vasiliades, L., Loukas, A., 2015. Multi-criteria analysis framework for potential flood prone areas mapping. Water Resour. Manag. 292, 399–418.
Parkin, J., Wardman, M., Page, M., 2007. Models of perceived cycling risk and route acceptability. Accid. Anal. Prev. 392, 364–371. http://dx.doi.org/10.1016/j.aap.
2006.08.007.
Pucher, J., Ralph, B., 2009. Integrating bicycling and public transport in North America. J. Public Transp. 12 (3).
Pucher, J., Buehler, R., 2008. Making cycling irresistible: lessons from the Netherlands, Denmark and Germany. Transp. Rev. 284, 495–528.
Pucher, J., Dijkstra, L., 2003. Promoting safe walking and cycling to improve public health: lessons from the Netherlands and Germany. Am. J. public health 939,
1509–1516.
Pucher, J., Dill, J., Handy, S., 2010. Infrastructure, programs, and policies to increase bicycling: an international review. Prev. Med. 50, S106–S125.
Rajabifard, A., Feeney, M., Williamson, I.P., 2003. Spatial Data Infrastructures: Concept, nature and GDI Hierarchy, chapter two; Developing Spatial Data
Infrastructures; from concept to reality. ISBN 0-415-30265-x.Taylor Francis.
Saaty, T.L., 1977. A scaling method for priorities in hierarchical structures. J. Math. Psychol. 153, 234–281.
Saaty, T.L., 2000. Fundamentals of decision making and priority theory with the analytic hierarchy process p. 458. RWS Publications. Retrieved from 〈http://cpe.njit.
edu/dlnotes/MIS645/MIS648_AFrameWorkfortheDevelopmentofDecisionSupportSystems.pdf〉.
Saplioglu, M., Uysal, M., 2015. Analysis and interpretation of headway data at unsignalized intersections. Eng. Sci. 9 (5), 1–20.
Schepers, J.P., Kroeze, P.A., Sweers, W., Wüst, J.C., 2011. Road factors and bicycle–motor vehicle crashes at unsignalized priority intersections. Accident Analysis.
Prevention 433, 853–861.
Schneider, R., 2005. TCRP Synthesis62: Integration of Bicycles and Transit, Washington, D.C.: Transportation Research Board of the National Academies, Published by
Transit Cooperative Research Program, Washington, D.C., United States.
Sener, I.N., Naveen, E., Chandra, R.B., 2009. An analysis of bicycle route choice preferences in Texas, US. Transportation 36 (5), 511–539. http://dx.doi.org/10.1007/
s11116-009-9201-4.
Shaheen, S., A., Cohen, Martin, E., 2013. Public bike sharing in North America: early operator understanding and emerging trends. Transp. Res. Rec. 2387 (83-92).
Silva, S., Alçada-Almeida, L., DIAS, L.C., 2014. Development of a web-based multi-criteria spatial decision support system for the assessment of environmental
sustainability of dairy farms. Comput. Electron. Agric. 108 (46–57). http://dx.doi.org/10.1016/j.compag.2014.06.009.
Stewart, K., McHale, A., 2014. Cycle lanes: their effect on driver passing distances in urban areas. Transport 293, 307–316.
Taylor, D., Mahmassasni, H., 1996. Analysis of stated preferences for intermodal bicycle-transit interfaces. Transp. J. Transp. Res. Board 1556, 86–95. http://dx.doi.
org/10.3141/1556-11.
Teschke, K., Harris, M.A., Reynolds, C.C.O., Winters, M., Babul, S., Chipman, M., Cripton, P.A., 2012. Route infrastructure and the risk of injuries to bicyclists: a case
cross over study. Am. J. Public Health 102, 2336–2343.
Titze, S., Stronegger, W.J., Janschitz, S., Oja, P., 2008. Association of built-environment, social-environment and personal factors with bicycling as a mode of
transportation among Austrian city dwellers. Prev. Med. 47, 252–259.
Turkish Min. of Pub. Works and Sett. Coun. of Urb, 2009. Turkish Ministry of Public Works and Settlement Council of Urbanization, Urban Technical Infrastructure and
Transportation Commission Report, Ankara.
Uz, E., 2003. Bisiklet Yollarının Geometrik Planlama Esasları ve Uygulaması In Turkish (Master thesis). Suleyman Demirel University, Isparta, Turkey.
Uz, V.E., Karasahin, M., 2004. Bicycle in urban transport. Turk. Eng. News J. 429, 41–46.
Vargas, L.G., 1990. An overview of the analytic hiyerarchy process and its applications. Eur. J. Oper. Res. 48, 4. http://dx.doi.org/10.1016/0377-22179090056-H.
Wang, G., Qin, L., Li, G., Chen, L., 2009. Landfill site selection usingspatialinformation technologyandAHP:acasestudyinbeijing, China. J. Environ. Manag. 908,
2414–2421.
Xu, Y., Sun, J., Zhang, J., Xu, Y., Wang, M., Liao, X., 2012. Combining AHP with GIS in synthetic evaluation of environmental suitability for China's 35 major cities. Int.
J. Geog Inform. Syst. 269, 1603–1629.
Yang, C., Wang, W., Shan, X., Jin, J., Lu, J., Li, Z., 2010. Effects of personal factors on bicycle commuting in developing countries: case study of Nanjing, China. Transp.
Res. Rec.: J. Transp. Res. Board 2193, 96–104.
Yuksel Project, 2001. Konya Büyükşehir Alanı Kentiçi ve Yakın Çevre Ulaşım Master Planı Çalışması, Bisiklet Ulaşımı Geliştirme Projeleri ve Konya Bisiklet Planı In
Turkish, 149p, Ankara, Turkey.

252

S-ar putea să vă placă și