Navigation – Plan du site
2019
920

Access Mobility to Local Railway Stations: Current Travel Practices And Forecast

Les pratiques de rabattement sur les gares ferroviaires: quelles évolutions possibles?
Sophie Hasiak

Résumés

Favoriser et accroître l’attractivité d’une offre ferroviaire invite à améliorer l’accessibilité aux gares. Cela amène à considérer la promotion des modes de transport respectueux de l’environnement et la mise en place d’une politique de gestion de l’offre de stationnement.
Connaître et comprendre les pratiques de rabattement actuelles sur les gares ferroviaires représente un enjeu important pour les collectivités pour être en mesure d’identifier des leviers d’actions pour agir sur les comportements de mobilité. Des travaux scientifiques internationaux se sont penchés sur ces déplacements de rabattement sur les gares ferroviaires en les étudiant en fonction des caractéristiques des individus, de la distance du trajet de rabattement, de la qualité de service de l’offre proposée en gare dont l’offre de stationnement et de l’environnement urbain autour de la gare. Mais ces travaux ne permettent pas de refléter les comportements d’accès aux gares des Français qui demeurent mal connus.
Cet article propose de qualifier le lien pouvant exister entre d'une part les pratiques de rabattement sur des gares françaises, le niveau de service de l’offre ferroviaire et d'autre part l’insertion spatiale de la gare dans le territoire. Cette approche s’appuie sur une évaluation de la part modale des modes de rabattement selon le profil d’une gare, notion combinant offre de transport et morphologie urbaine. Elle propose aux collectivités locales une approche méthodologique d'estimation des parts modales pour les aider dans les décisions à prendre sur les enjeux d’accessibilité aux gares et leurs aménagements sans avoir besoin de recourir à la mise en œuvre d’enquêtes territorialisées.

Haut de page

Texte intégral

Introduction

1Metropolitan areas face challenges due to some complex issues of their territories’ structuring. The process of metropolisation has led to a strong polarization of jobs and services in major urban centres (Carpentier, 2010). Moreover, technical progress in transportation and road investment have improved conditions of car travel (Brons et al., 2009; Carpentier, 2010; Dupuy, 2010; Wiel, 1996, 1999), and have maintained a stable travel time budget for individuals (Zahavi, 1973). They have thus contributed to improving accessibility of less dense urban areas and to the opening up of "broader areas to urbanization" while still being "linked in various ways to the city centre without living there" (Wiel, 1999). Transport policies over the last decades have thus participated in amplifying the processes of urban growth, urban sprawl and periurbanisation, generating car dependency for local residents (Dupuy, 1999; Motte-Baumvol et al., 2012; Aguilera et al., 2017).

2Metropolitan areas have therefore to act in response to land development issues, which cover both impacts of urban sprawl and mobility, and environmental consequences (noise pollution, road congestion, air pollution). Consequently, public policies focus on ways to tackle the negative effects of these processes by seeking a better consistency between shape of cities and "forms of mobility that may be considered desirable" (Le Néchet, 2011). The opportunity to serve these territories by rail may be a lever for action, indeed the rail enhancement has been considered as an important challenge of land planning policies for several years. It takes shape according to the principle of Transit Oriented Development (TOD), an urban concept relying on increasing density of buildings on both sides of railway lines (metro or train stations), and promoting access to public transport networks (Gallez et al., 2015) by walking, cycling or using public transport.

3Thus, the policy of the railway system enhancement is considered as a way of acting on challenges related to attractiveness and territorial sustainability (Delage, 2013). It is also potentially a response to accessibility issues to metropolitan amenities (Richer, 2012). It is built around a dual challenge related to the interrelation between urbanism and mobility issues. On the one hand, it relies on the implementation of a rail-oriented urban planning policy based on urban densification around railway stations. On the other hand, it concerns the promotion of rail travel on offer (frequency, train comfort, regularity, information, cost, etc.). Moreover, the train travel is a link of a multimodal journey chain, which includes access to and from the railway station. As a result, accessibility to the rail station is a key factor of rail mobility. It corresponds to a challenge of increasing rail use (Givoni et al., 2007; Brons et al., 2009). As a consequence, promoting accessibility to stations encourages public policies to analyze how easy it is for people to reach a rail station according to their location and the transport on offer (Bavoux and Beaucire, 2005). It is then a question of evaluating spatial accessibility to the station defined by its capacity to be reached by people with the means of transport they wish to use from their geographical position. Besides, accessibility to the rail network is a deciding factor in the choice of the train (Brons et al., 2009). The quality of routes to access to and egress from stations has an influence and determines their attractiveness (Krygsman et al., 2004; Hasiak et al., 2016a). Providing good accessibility to railway stations is important to increase the use of the train (Brons et al. 2009; Keijer et al., 2000) and the catchment area of public transport (Krygsman, 2004; Debrezion et al., 2009), in the way that it strengthens its attractiveness.

4Therefore, knowledge of the real conditions of accessibility is a key element to be considered in a policy of territorial enhancement. Therefore, local authorities in charge of transport and land planning policies have to understand how people reach railway stations, and the modes they use (Halldorsdottir et al., 2017) to be able to act on both railway station accessibility and railway supply.

5The literature underlines the importance of international studies and research assessing mobility practices to reach railway stations by collecting field data or by implementing modelling tools (Cervero et al., 1995; Debrezion et al., 2009; Department for Transport (UK), 2012; Rak et al., 2014; Creemers et al., 2015; Halldorsdottir et al., 2017). It also points out that in France, knowledge of access modes to railway stations mainly comes from local studies only based on field surveys (Audiar, 2013; Cerema Nord Picardie, 2016, 2018; Cete Nord Picardie, 2012; L'hostis et al., 2013; Mamoghli, 2009). They provide a picture of mobility practices observed for a given station, but do not allow us to generalize the results for all kinds of stations. By default, modal shares of access travels are assessed according to some expert opinions (Aguram, 2018). All the outcomes show that there is a variability of access modal shares according to types of stations.

6Given the large number of rail stations covering France, stakeholders have a growing interest in "central stations" located in metropolises. Indeed, since the 2000s, metropolises are considered as strategic areas to undertake important land planning projects in and around train stations (Roudier, 2015; Delage, 2013). These rail stations are their investments priorities, both in terms of studies and maintenance work. For the other stations, they may carry out some studies on their possible development into intermodal hubs but it remains marginal. Indeed, conducting studies such as field surveys to understand how people get to a station is an important cost for authorities. Therefore, questioning accessibility to rail stations invites the scientific community to look into potential methods at a lower cost. This would provide framing data on travel practices to reach a rail station.

7Furthermore, international research mentioned above has mainly analyzed trips to stations according to users' characteristics, access distances, level of transport services including parking offer, or station environment. Considering this knowledge, this paper aims to define the potential link between access modes of transport, the transport profile of stations and their spatial environment characteristics all together. It also wants to propose a method, which links the spatial characteristics of stations (transport and environment) with access modes, independently of users' characteristics.

Literature review and issues

8Research and studies focusing on access and egress modes to train stations mainly rely on two main methodological approaches. The first one is based on descriptive analyses of access and egress travel behaviors to get to a station. Using data collected from train users for a specific station, these analyses emphasize the main modes used to access a station (and also the mode used to egress from the destination station). Some cross-analyses may be carried out to link modal practices with users' characteristics or with the urban morphology of places of residence or station areas (NSW government, 2006; Cerema Nord Picardie, 2016, 2018; Cete Nord Picardie, 2012). But it only provides a picture of access mobility practices observed for a specific station, without the opportunity to generalize the results for all kinds of stations.

9The second type of methodology concerns modeling approaches, which aim to highlight determinants of choice of access and egress modes by using utility maximization functions (Brons et al., 2009; Bergman et al., 2011; Cervero, 2002; Chakour et al., 2014; Debrezion et al., 2009). The main objective of modeling methods is to measure the influence of different variables on the choice of access or egress mode. Overall, this international work tends to conclude that this modal choice may be summarized by a function including three items: the service level of transport supply, the features of the station environment and the characteristics of train users.

10This work does not allow us to identify a convergence of the conclusions on access and egress modal choice criteria. Admittedly, there is a consensus to say that access distance to reach a station is the most important factor of access modal choice (Chakour et al., 2014; Givoni et al., 2007; Halldorsdottir et al., 2017; Keijer et al., 2000; Krygsman et al., 2004). As expected, the probability to walk to a station decreases when the distance to reach the station increases. Thus, Debrezion et al. (2009) prove that, in the Netherlands, walking is the most probable mode choice for a short distance, range of up to 1.1 km, and cycling is most used in the range of distance between 1.1 and 4.2 km. In France, according to observed practices, the rule used by engineering firms is to consider the threshold of 0.8 or 1 km for the walk and 3 km for the bike. Daniels et al. (2011) talk about guidelines using the value of 400 m and its multiples as a reference value according to the type of stations.

11However, we can’t talk about a consensus for other modal choice determinants as the results appear to be rather contrasted. Indeed, in some cases, parking capacity and availability are two significant factors that influence the choice of the car (Cervero et al., 1995; Debrezion et al., 2007; Bergman et al., 2011). In other cases, especially for territories in which the modal share of walking to stations is high, parking availability has no effect (Halldorsdottir et al., 2017).

12Besides, some modeling results highlight some correlations between travelers' characteristics and their modal choice to reach a station (Bergman et al., 2011; Chakour et al., 2014). The main features concern the level of car ownership and the occupation status (Halldorsdottir et al., 2017). Car ownership has a positive effect on choosing the car to access a station (Creemers et al., 2015). Car availability may also influence access mode choice (Creemers et al., 2015). However, it does not have a strong effect on choice in territories where people prefer to use green modes (Givoni et al., 2006, 2007; Debrezion et al., 2009). The influence of socio-demographic factors (gender, income, age) varies according to the type of transit mode (Creemers et al., 2015). There is no convergence of outcomes and some researchers (Krygsman et al., 2004; Cervero, 2001) underline a little influence of variables related to gender, age, income and car availability on feeder modal choice to rail stations.

13However, it seems that there is more consensus on the impact of the station environment on access mode choice (Krygsman et al., 2004; Chakour et al., 2014). In particular, density is said to have a significant influence on rail access travel (Bergman et al., 2011; Cervero et al., 1995). Hence, the higher the density, the more people are likely to walk. Some work points out the effect of the station surroundings morphology (local built environment, population density and diversity) on the walking distance (Cervero, 2002; Bergman et al., 2011). According to the evaluation of the average walking distance, the value of 400 m, or even 800 m is considered as a key distance, as a "rule of thumb" (Daniels et al., 2011). Unfortunately, there is a discordance about using a mean value. Indeed, some research underlines differences in walking behaviors according to the geographical location of the station (Corpuz et al., 2005; Chakour et al., 2014). And it appears that people who live in highly urbanized areas are likely to walk more (Cervero et al., 1995) while suburban commuters are more dependent on non-walk modes for train access (Bergman et al., 2011).

14As a conclusion, there has been a great deal of work to understand and explain the choice of access mode to railway stations based on criteria of distance, time and supply (Givoni et al., 2007; TCRP, 2012), on features of the station environment or even on people characteristics. They bring a shared knowledge to the scientific community. But they mainly concern countries outside France. Indeed, in France, knowledge of access modes to railway stations mainly results from few local studies carried out on major stations (see previous section). The transposition of this international knowledge to the French situation does not seem easy, as each country has different spatial and behavioral characteristics. Therefore, we have chosen to conduct an operational research to help French local authorities to define their policy of railway system enhancement. Indeed, as it is the case in several countries all over the world, authorities have public financial constraints, which can be a major barrier to the implementation of data collection surveys in order to know how people access stations. Our research aims to provide a simplified method, which would estimate the modal shares of access modes according to the type of railway stations.

15Given the main learnings highlighted by international work mentioned above, we assume that it is possible to define a model of station access modes according to the station profile. By railway station profile, we mean both the characteristics of railway supply and the station environment.

16After specifying our methodological choices, the paper will first characterize access travel to French stations and will propose a brief international comparison of these practices. Then, it will analyze the different profiles of French local stations using a specific typology. After focusing on people profiles according to train stations, it will evaluate the travel practices to reach stations according to their typological profile. Different scenarios of mobility policies, especially trend and voluntarist ones, based on the promotion of walking and cycling, will be considered in the assessment. Finally, the paper will open the discussion on the challenges of improving accessibility to stations and the necessity of working on communication and awareness campaigns adapted both to the stakes of each station and to sustainable mobility.

Methodology

The field study

Access travel to a railway station

17Our research focuses on feeder travels to stations by train users who have chosen the train as their main mode. In this paper, by feeder travels, we only mean trips made to reach the train station from home (figure 1). First of all, French studies (Cerema Nord Picardie, 2016, 2018) have underlined that a train trip essentially refers to a multimodal transport chain, which starts at home and which concerns a single pattern, such as work, shopping or leisure. About 90 % of train users travel from or to home.

Figure 1: Home to work train travel: a segment of a multimodal transport chain

Figure 1: Home to work train travel: a segment of a multimodal transport chain

Conception S. Hasiak

18Then, egress modes differ from access modes used in periurban areas. There is an asymmetry of modes used for access and egress. In particular, people would rather use public transport or walk to reach their final destination, once they are in the urban city (Givoni et al., 2007; Cete Nord Picardie, 2012; Creemers et al., 2015). This difference mainly results from the urban morphology of the origin (less density in periurban areas) compared with that in the destination (higher density in urban centres).

19Considering these elements, we chose not to take into account both access travels to train from a location different from home and egress travels from railway stations.

Local railway stations

20In the first section, we have pointed out that French metropolitan train stations focus a growing interest of stakeholders, since they are considered as strategic areas for land planning projects (Roudier, 2015). The situation is quite different on the other types of stations, even if stakeholders are aware of their importance in territorial policies. It then seems relevant to focus on train stations located outside metropolises to define a framework methodology on evaluating access travel to railway stations without the need to carry out a field investigation.

21These stations may be located in periurban areas, which surround most urban centres. The morphology of these periurban areas is partly urban and partly rural, and there is a large majority of residents going to the urban centre to work, to go shopping or for amenities (Chapuis, 1995). The stations may also be located in small and medium-sized cities that have "an intermediate status between metropolises and small local urban centres" (Santamaria, 2012), and keep themselves away from the metropolisation process (Commerçon, 1999).

22Moreover, considering the specificities of the Paris Region (density of population and housing, high road congestion, polarisation of jobs), which includes the Capitale (Paris) and its surroundings; we chose to only consider stations situated outside this region.

23In the following of this paper, we will use the expression "local train stations" to talk about these stations. These local train stations have some specific features. Both rail traffic and passengers’ flows are lower than in main stations located in metropolises. Moreover, they are only served by regional train services, that is to say railway lines connecting cities located in the same administrative region or in the surrounding one. These regional train services differ from the other services operated on long distance interurban connections between major urban centres (either on the high speed network or the conventional one) (see table1).

Table 1: Segmentation of the French train services

Type of train services

Competent authority

Rail operator

Trademark

Long distance services

High speed services (running on a specific rail network)

SNCF

SNCF

"TGV Inoui"

Interurban services (running on the conventional railway network)

The French State

SNCF

"Intercités"

Medium distance services

Regional services

Region

SNCF

"TER"

Conception S. Hasiak

24We used two types of database to select the local rail stations as defined previously. The first one is the territorial breakdown of urban areas defined by the French national institute for statistics and economic research (INSEE) in 2011. It provides a spatial classification of the different spatial entities, distinguishing main urban areas, other urban areas, and remote cities (see figure 2). An urban area includes the urban centre and the periurban suburbs defined by the proportion of working people (value of 40 %) who go to work in the urban centre. This classification provides a specific view of the territorial influence of French cities.

Figure 2: Classification of French urban areas

Figure 2: Classification of French urban areas

Source: INSEE 2010; conception: S. Hasiak

25The second database relies on the segmentation of French rail stations defined by SNCF, the historical rail operator. This base is used by SNCF to set pricing policy of its railway stations uses (SNCF is the owner and the manager of French railway stations). It is based on both type of train services and attendance level. The local stations considered in this work will correspond to some stations which have a tarification "b" and "c" (see table 2).

26The crosschecking of these databases provides a first file of local train stations.

Table 2: Diversity of French railway stations

Type of stations (classification SNCF)

Tariff classification

Attendance threshold (F) - Pass./year

Number (2016)

National railway stations

Type "a"

F> 250,000

122

Regional railway stations

Type "b"

100,000 <F< 250,000

956

Local railway stations

Type "c"

F<100,000

1,890

Source: SNCF open data (data.sncf.com) 2018

Database combining several households travel surveys

27In France, there is no specific railway mobility surveys, compared to other countries (national travel surveys in the UK, in Australia (Department For Transport, 2012; NSW government, 2006). We evaluate French railway mobility on the basis of households travel surveys (HTS) carried out on wider perimeters of urban areas. These surveys based on a national methodology provide data on weekday travels made the day before the survey day, by collecting the travel chain from the origin to the destination.

28Since a HTS survey concerns all types of travel modes and considering the number of train travels in a single survey compared to the number of car, public transport and foot travels, we chose to work on a corpus of several survey results that have been gathered together to have a more significant sample of train travels. It corresponds to 25 households travels surveys carried out on different French urban areas and regions between 2009 and 2016, outside the region of Paris (figure 3).

Figure 3: The French Households Travel Surveys (HTS) used

Figure 3: The French Households Travel Surveys (HTS) used

Conception Cerema

The selection of the "local train stations" sample

29We have extracted train journeys among travel data of the 25 HTS surveys. After having filtered rail stations and type of travels (only travels from home are considered), our sample includes over 2,500 train journeys. This number provides a variety of feeder travel behaviors to railway stations.

30The main characteristics of train users are summarized in table 3. There are currently as many men as women taking the train for their daily travel. Young adults and working-age people mainly represent the customer of local trains. Indeed, the train is mainly used for home to work or study travels.

Table 3: The main characteristics of passengers of local railway stations

Age group

%

Journey purpose

%

Occupation status

%

under 18

12.5

work

61.3

Executives, higher intellectual occupations

20.4

18-25

20

study

25.2

Intermediate professions

17.9

26-45

34.2

purchases

2.8

employees

27.3

46-65

30.8

others

10.7

workers

5.0

Over 65

2.6

Pupils, students

27.6

Inactive people

0.9

others

0.9

Source: data of 25 Households Travel Surveys

31Moreover, 71 % of train users have a driving licence. If we exclude teenagers who are in an accompanied driving process, 23.9 % are captive train users. Among the population of users, 57 % of them have an alternative to train trips (they have both the driving licence and a car) but they yet chose the train.

32These train journeys correspond to 567 local rail stations located throughout France but outside Paris region. In France, there are nearly 3,000 stations, including 2,600 stations outside Paris region. If we exclude the 112 main stations outside Paris, our sample corresponds to 23 % of small and medium-sized French stations (table 4).

Table 4: Diversity of French railway stations outside the Paris region

Daily attendance Median (2016)

Number (2016)

Number in our sample

Percentage of the total number

National railway stations

5,900 pass./day

112

0

0 %

Regional railway stations

939 pass./day

613

322

52 %

Local railway stations

53 pass./day

1,884

245

13 %

Source: SNCF open data

33The French railway network includes numerous stations called "local railway stations" by SNCF (72 %), which correspond to the tarification segmentation "c". The sample only comprises 13 % of these type of stations. This percentage may be explained by the fact that these stations do not have a high level of attendance. Indeed, the attendance median is close to 53 passengers/day, which corresponds to a low level. Besides, there are less than 13 passengers per day in 25 % of these stations (471 stations). Therefore, the likelihood of having surveyed an individual taking the train in a low-traffic station is low.

Figure 4: Our sample of French local train stations

Figure 4: Our sample of French local train stations

Conception Cerema

34The sample of local rail stations concerns the main surroundings of French metropolitan areas (figure 4). That contributes to provide a sufficient representativeness of the French situation. The table 5 summarises the main features of this sample of local railway stations, which are located in less dense and residential areas.

Table 5: Characteristics of cities (administrative boundaries) in which local stations are geographically located

N =567

Population density (inhab/ha)

Housing density (per ha)

Proportion of built-up area (%)

Proportion of collective dwellings (%)

Minimum value

0.22

0.12

1.3

0.0

First quartile

2.00

0.97

12.7

20.0

Median

4.57

2.34

22.9

37.6

Third quartile

9.50

4.91

38.9

57.1

Maximal value

90.94

53.19

99.1

95.6

Source: INSEE data 2017; Cerema, DGFIP data 2017

35The median of trains frequency corresponds to six trains during morning peak hours (figure 5 a, b), that is to say a train every 20 minutes, and to ten trains in off-peak hours (equivalent to a train every 42 min). 25 % of our stations have a train frequency higher than one train every 13 minutes at peak hours. Besides, the median of attendance is close to 550 passengers per day (figure 6), the first quartile corresponds to 175 passengers per day, and the third one to 1,350 train users per day.

Figure 5: a) frequency of trains at peak hours

Figure 5: a) frequency of trains at peak hours

Source: SNCF data 2017

Figure 5: b) at off-peak hours

Figure 5: b) at off-peak hours

Source: SNCF data 2017

Figure 6: Number of train passengers per day

Figure 6: Number of train passengers per day

Source: SNCF data 2017

The different methodological steps

36First, the implementation of a reference set of indicators characterizing stations profiles (mentioned above) required to change travel data (one line corresponds to one train travel made by one user) into "station data" (one line corresponds to one rail station).

37Then, we used both descriptive statistical and econometric methods. The descriptive statistical analysis provides some findings about access travel practices to French local train stations. Then, considering the profile of railway stations based on criteria of transport supply and station environment invited us to implement a typology. We used a principal component factor analysis (PCA) to identify the correlated factors among the reference set. This method also highlighted the criteria that most differentiate train stations (factorial axes). The rotation phase of this PCA was not led. An ascending hierarchical classification helped us to classify the stations according to these factors. It refers to the Ward's hierarchical procedure, which aims to minimize intra-class inertia and to maximize inter-class inertia to obtain the most homogeneous classes as possible. A specific next section will provide the main characteristics of this process.

38Our approach also involved to check a working hypothese. As we assumed that we could assess modal travel shares to access stations as a function of stations profile, we had to evaluate the variability of the profile of train users according to the profile of stations. Therefore, we used an independence t-tests method (chi-square tests).

39Finally, we built some scenarios of mobility policies to assess the possible range of evolution of the current modal shares observed. We chose to consider the feeder access distance, considering that there are some virtuous travel practices, which may be encouraged for short trips. The distance threshold for such short trips was appraised from the distribution of travel practices observed in our sample (the case of walk distance) or on data commonly used (the case of bicycle journeys).

Main results

Travel practices to access French local railway stations

Feeder distances

40Overall, for all the stations of our sample, the maximum distance to reach a station is evaluated around 7.9 km according to the 90th percentile, whatever the mode (table 6). By comparison, considering French railway operator data on national stations, this value is around 20 to 30 km for major French railway stations. In Hasiak et al. (2016a), the variation of access distance for regional stations is in the range from 4 to more than 8.5 km, value appraised on the basis of 99 stations. Our result is quite faithful with this conclusion.

Table 6: Access distance to local railway stations

N = 2,520

mean value

Standard deviation

10th percentile

20th percentile

50th percentile

80th percentile

90th percentile

Access distance (meters)

3537

12,810

333

667

1,540

4,870

7,940

Source: data of 25 Household Tavel Surveys

41However, this maximum access distance to station hides some disparities. Indeed, 80 % of access travels to station are less than 4.9 km. Besides, there are many small travels. Nearly 50 % of access travels correspond to a route of less than 1.5 km (table 6). On the one hand, these results underline that these local stations can be referred as “proximity stations”, which means that they have a strong influence on the nearest environment around the station. On the other hand, we can’t deny that their influence goes beyond and attracts people who live farther. They also have a function of “feeder stations”.

Modes used

42In the panel, considering only travels from home, the car is the first mode used to reach a local station, with a modal share of 53 %. And 37 % of access travels concern driving car trips. Walking ranks second with 36 % of train users who walk to stations. The use of public transport is quite limited and less than 10 %. This result may be partly explained by the location of stations in periurban and rural areas, which are well less served by public transport networks. And cycling to station still remains a marginal mode to access this type of stations in France (figure 7).

43By comparison, for the national stations located in dense urban areas and metropolises (and outside Paris region), the travel practices are very different since walking is the first mode used and public transport the second one. The car is relegated to the third position with a modal share less than 20 %.

Figure 7: Access modes to French regional stations (all travel purposes)

Figure 7: Access modes to French regional stations (all travel purposes)

Sources: open data SNCF 2017 and HTS surveys

44We tested the influence of the geographic location of these local railway stations on modal access pratices to reach them. These latter were divided into two classes, the North and the South of France, according to a virtual separation dividing the geographic area in two equal parts. Whatever the geographic location, there are the same modal behaviors to access a local train station (figure 8). Walking is the first mode used in the same proportions, and driving the second.

Figure 8: Access modes to French local railway stations according to their geographic location

Figure 8: Access modes to French local railway stations according to their geographic location

Source: HTS surveys

45Some international work underlines that distance affects the modal choice (Keijer et al., 2000; Givoni et al., 2007). This finding is also the same in our case. Indeed, walking is the predominant mode when the distance is under 1.5 km and over this value, the train user will massively choose the car to reach the railway station (figure 9).

Figure 9: Modal shares to reach a local station according to access distance (in km)

Figure 9: Modal shares to reach a local station according to access distance (in km)

Source: HTS surveys

46The comparison of these results with travel behaviors in other countries highlights some differences, linked both to mobility policies and to mobility culture (table 7). Hence, walking is the primary access mode for trips from home to railway stations in Denmark and in Australia. Cycling is the first feeder mode in the Netherlands, which is known as a country where the use of bicycle is common and amongst the highest in the world (Givoni et al., 2007). While the car is only the fourth most popular mode used in the Netherlands, driving to access suburban stations still remains the first mode used in Montreal region, in Slovenia, in North America and also in France. Finally, the average distance to station is quite similar in the Netherlands and in France (3.5 km).

Table 7: International comparison of modal shares of access modes to stations

Country

Denmark

Australia

Netherlands

Canada

Slovenia

continent

Europe

Oceania

Europe

North America

Europe

Stations type

Suburban stations of Copenhagen

Sydney area stations

All stations

Suburban stations (Montreal Region)

Suburban stations of Podravje Region

Walking

51 %

50 %

20 %

25 %

26 %

Cycling

24 %

2 %

38 %

1 %

5 %

Car driver

5 %

17 %

7 %

51 %

60 %

Car passenger

5 %

17 %

7 %

12 %

Public transport

15 %

14 %

27 %

11 %

8 %

Average feeder distance

Not available

Not available

km

4.9 km

4.9 km

Sources: data from scientific papers of Chakour (2014), Daniels (2011), Givoni (2007), Halldorsdottir (2016), Keijer (2000) and Rak (2014)

Train access behaviors

47The average values of access travel practices presented previously hide some disparities according to the profile of users. Gender has a weak influence on how to get to a station (figure 10), excepted the fact that women would be more likely to be dropped off at a train station or to take public transport or that there are fewer women who cycle. Nethertheless, these results do not show an important difference of behaviors, which is confirmed by the work of Givoni (2007).

Figure 10: access modes to reach a local station according to gender

Figure 10: access modes to reach a local station according to gender

Source: HTS surveys

48By contrast, age has an effect on the mode used to access a station. Young people under the age of 25 walk or are dropped off by car, with a modal share quite similar (figure 11). When people are in working age or do work (from 26 to 65 years old), the situation is different since the car is the first mode used to go to the station, especially in its component of car driving. Elderly people have a similar behavior to young people, except that they are the driver of the car. These results highlight an increased car use with age, which was also underlined by Givoni (2007).

Figure 11: access modes to local stations according to the age of train users

Figure 11: access modes to local stations according to the age of train users

Source: HTS surveys

49Moreover, we observe a similarity of access travel practices according to the travel purpose (figure 12). The correlation between the age and purpose variables may explain these results. Thus, train users going to work will prefer to take the car as a driver to reach a station, while pupils and students will walk or be dropped off. For all other purposes (shopping, visit, administrative procedure...), walking is the main mode used to access a station.

Figure 12: access modes to local stations according to purposes

Figure 12: access modes to local stations according to purposes

Source: HTS surveys

50Besides, if we consider the train users who have voluntarily made the choice of the train even though they would have the opportunity to drive (57 % of the panel), these latter however prefer the use of the car to access a station (proportion of 70 %). Walking is the second choice of access mode for 22 % of them.

51Finally, the breakdown of access modes according to departure time shows a decline in the use of the car with a late departure time (figure 13). Walking then becomes the first mode of access for a travel starting after 9 am. The availability of a parking space may be one of the explanatory factors of this phenomenon.

Figure 13: access modes to reach a local station according to departure time

Figure 13: access modes to reach a local station according to departure time

Source: HTS surveys

Typology of local stations

Definition of a station profile: setting up a reference set of indicators

52Considering research learnings about rail stations attractiveness underlined previously, we defined the profile of a railway station according to two topics: train service offer at the station and station environment. First, the railway transport offer is characterized by criteria related to train frequency during two time periods: the morning rush hours (7 am – 9 am) since we analyse access to stations, and the off-peak hours corresponding to the period between 9 am and 4 pm.

53We decided not to take the parking offer at the station into account for two main reasons. The first one relates to data availability: there is a lack of exhaustive database of parking offer in France. The second reason rests on the fact that parking pressure on these local stations, located outside major urban poles, is much lower than on urban stations. There is no particular difficulty to find a parking space, whether on the station park or in the street.

54That may explain that car parks of stations located in the outskirts of cities are often free of charge (Habib et al., 2013). Indeed, introducing parking fares is considered as a measure to manage capacity limitations (Habib et al., 2013) by controlling the number of train users who reach the station by car. La Paix Puello et al. (2014) show that parking costs have an negative effect on choosing the car to get to a station. Besides, both the parking spaces remaining and the parking fee have a significant effect on the choice of the railway station where people take their train (Chen et al., 2015). Car park pricing is thus mainly implemented in central train stations in metropolises due to a heavy land pressure and restrictive parking policies. Besides, our sample includes less than five stations located near major urban cities, which have their car park controlled by an access control system: only train users having a rail pass are allowed to park their car in the station parking. In this context, pricing is not considered among our criteria. Moreover, we may mention that pricing policy of French car parks located at railway stations is not well known. It requires collecting the information for each station by exploring each website.

55Then, according to Cervero et al. (2002), indicators of territorial density characterize the station environment. Only density and diversity are considered since the dimension of "design" suggested by Cervero is complex to implement on a large sample. The built environment is evaluated in terms of density of population, density of housing, housing typology and built-up areas including built land, roads, railways, car parks…

56The indicators toolkit characterizing train stations profiles is thus composed of a set of seven indicators summarized in figure 14.

Figure 14: The set of indicators characterizing the profile of a train station

Figure 14: The set of indicators characterizing the profile of a train station

Conception S. Hasiak

Main results of the PCA process

57Based on this framework, the Principal Component factor Analysis (PCA) led on the reference set of indicators identifies two main factorial axes, which explain 84 % of the total inertia of the point cloud (table 8).

Table 8: Principal Component factor Analysis (PCA) – eigenvalues of the correlation matrix

Factorial axis

eigenvalue

difference

proportion

Cumulative proportion

1

3.49392

1.96309

0.5823

0.5823

2

1.53082

0.95464

0.2551

0.8375

3

0.57618

0.37177

0.0960

0.9335

4

0.20441

0.05986

0.0341

0.9676

5

0.14455

0.09442

0.0241

0.9916

6

0.05013

 

0.0084

1.0000

58The territorial density indicators contribute to define the first axis whereas the second axis is more stuctured by train supply factors (see table 9).

Table 9: Principal Component factor Analysis (PCA) – eigenvectors

Attributes

Axis 1

Axis 2

Axis 3

Axis 4

Axis 5

Axis 6

Population density

0.4754

-0.2972

-0.2028

-0.3433

0.1028

-0.7184

Housing density

0.4734

-0.3068

-0.1141

-0.4517

-0.0727

0.6778

Built-up area proportion

0.4614

-0.2425

-0.2230

0.8173

-0.0792

0.0667

Collective dwellings proportion

0.3823

0.0480

0.9167

0.0594

0.0759

-0.0432

Train frequency at off-peak hours

0.3101

0.6178

-0.1019

-0.0811

-0.7058

-0.0838

Train frequency at morning peak hours

0.3079

0.6121

-0.2130

0.0079

0.6885

0.1053

59The hierarchical ascending classification process (Ward's method) leads to identify four clusters (figure 15).

Figure 15: Cubic Clustering Criterion (CCC) according to the number of local railway stations clusters (NCL)

Figure 15: Cubic Clustering Criterion (CCC) according to the number of local railway stations clusters (NCL)

Clustering of stations profiles

60The principal component analysis combined with the ascending hierarchical method contribute to define a typology of local stations according to their profile based on four clusters.

61The main characteristics of stations for each cluster are given in table 10.

Table 10: average value of indicators according to the cluster of stations

cluster

1

2

3

4

average

Std*

average

Std*

average

Std*

average

Std*

Population density

3.531

2.859

8.949

6.545

20.345

10.492

60.643

18.761

Housing density

1.724

1.445

4.856

3.730

11.555

6.385

36.277

10.755

Built-up area proportion

0.186

0.120

0.323

0.157

0.616

0.160

0.906

0.070

Collective dwellings proportion

0.290

0.190

0.551

0.191

0.579

0.173

0.788

0.181

Train frequency at off-peak hours

8.442

4.625

24.874

9.954

11.360

5.205

20.200

8.270

Train frequency at morning peak hours

5.315

2.301

14.027

4.487

6.888

2.952

11.100

4.095

Std*: standard deviation

62Then, based on this typology, we evaluated the influence area of each group of stations by calculating the average feeder distance. The figure 16 presents a schematic representation of this typology combining the two steps.

63A first cluster (see cluster 1 on figure 16) gathers local stations located on low-density areas, mainly in periurban areas. We note a predominance of individual and isolated housing. Population density is lower than four inhabitants per hectare. The service level of railway supply is quite average: there is a good supply in peak hours (one train every 22 min) but a degraded supply during off-peak hours (one train every 50 min). The second cluster (see cluster 2 on figure 16) also corresponds to local stations located on low-density territories. However, there are two major differences distinguishing it from the first group. On the one hand, housing structure is mixed with individual and collective dwellings. On the other hand, the train service supply has a good and even a very good level with a train every 9 minutes on morning. In addition, frequency during off-peak hours is higher than peak hour frequency of the first cluster. Then, the third profile (see cluster 3 on figure 16) refers to railway stations located in denser areas than the two previous groups, both in terms of built-on spaces, population and housing. 90 % of these stations are located in the major urban areas. However, the train supply level is quite intermediate compared to the second cluster. Finally, the last group (see cluster 4 on figure 16) includes few stations located in the most urbanized areas of our sample, with the higher population and urban density. Housing is essentially composed of collective dwellings. And the train supply is very good and is equivalent to the one of the second cluster.

Figure 16: Typology of station profiles, clustering of stations sample and statistical indicators

Figure 16: Typology of station profiles, clustering of stations sample and statistical indicators

Conception S. Hasiak

64The influence areas of these four station profiles are quite equivalent. Indeed, the maximum access distance (9th decile of access travel distances) is between 7.4 and 8.4 km (figure 17). This result is consistent with the conclusion of Hasiak et al. (2016a), which emphasize that the influence area scope depends on the level of train service and a high level may contribute to extend the average distance value of the attractivity area.

Figure 17: Distribution of access distances according to the station profile

Figure 17: Distribution of access distances according to the station profile

Source: HTS surveys

65This threshold of access distance varies according to the mode (figure 18). People are willing to walk up to 1.3 km whatever the profile of station they have to reach. On the contrary, the driving distance to get a regional station varies, depending whether you are a driver or a passenger and depending on the train station. A driver may travel a longer distance to access a station than a passenger will do. Lastly, the relevance of a feeder travel by public transport is about a trip less than 4.3 km.

Figure 18: Distribution of access distances according to access modes

Figure 18: Distribution of access distances according to access modes

Source: data from HTS surveys

Is the profile of train users the same according to the station profile?

66Access travels to stations are related to people who have made their modal choice according to their own characteristics and point of view. Therefore, we had to verify the working hypothesis of our research, that is to say the potential difference in profiles of train users according to station profiles. We carried out a Chi-square test to measure the possible independence between the characteristics of individuals and the station profile typology (table 11).

Table 11: Dependence test of qualitative variables

Chi-square (p - value)

meaning

Main occupation status x station type (4 clusters)

53.7978 (<.0001)

Rejection of the hypothesis of independence

Main occupation status x station type (clusters 2-3)

0.4782 (0.9237)

Hypothesis of independence not rejected

67Statistically speaking, it appears that train users characteristics are not strictly identical according to station profile. In particular, the structure of train users who reach a station belonging to the cluster 1 seems to differ from the other groups (table 12). Indeed, there are an over-representation of schoolchildren (proportion of 20  % compared to 10  % in average) and an under-representation of working people (59  % against 66  % in average). Conversely, if we exclude the fourth group, which is not significant because of its weak sample, the main occupation status of train users belonging to the second and third groups is very similar.

Table 12: Main occupation status of train users according to the profile of station

Cluster 1

Cluster 2

Cluster 3

Cluster 4*

Working population (%)

59

67

66

64

Students (%)

13

14

15

15

Pupils (%)

20

10

11

7

Other population (%)

8

9

8

15

* small sample

68Moreover, there are also some differences according to gender and age (table 13). This result is coherent since these variables are correlated. While the proportion of men is very similar for the first and second groups, women users are more important in the third group. As well, young people are less in the groups 2 and 3 but there are more people between 26 and 65.

69Besides, the results emphasize a net difference of car availability proportion according to clusters (table 13). The percentage of captive train users with no car availability is more important for the third group (the fourth group is not representative). On the contrary, the proportion of people who are in a situation of modal choice (daily availability of car) is higher for the cluster 2.

Table 13: Other socio-demographic features of train users according to the profile of station

Cluster 1

Cluster 2

Cluster 3

Cluster 4*

Men/women (%)

50 / 50

49 / 51

53 / 47

44 / 56

Under 18 years old (%)

16

9

10

10

18-25 years (%)

20

20

25

25

26-65 years (%)

61

68

63

60

More than 65 years old (%)

3

3

2

5

No car availability (no car or no driver licence) (%)

10

9

13

25

Daily availability of the car (%)

53

61

55

51

Some constraints of car availability (%)

37

29

32

24

* small sample

70In conclusion, we can not say that the profile of train users is similar whatever the stations. Although we are aware of this potential influence, and considering the challenge of this research aiming to identify an easy method to local authorities, we will not consider the socio-economic factor in the evaluation of modal access shares according to the station profile. Indeed, assessing the proportion of people according to their main status, their gender, their age or their car availability implies to have geo-referenced data for the influence area of stations. This type of data requires specific requests to the national statistical institute.

71As highlighted by scientific work (Corpuz, 2007; De Witte et al., 2013), socio-demographic characteristics of people influence the modal choice. The analysis of access modes to local stations according to the socio-demographic structure of train users underlines some differences. In France, daily train users mainly correspond to working people, students and schoolchildren. In addition, they have different practices to reach a railway station. Thus, working people prefer the car (59 % of observed access travels), then they are willing to walk (32 % of modal share). Students also use the car (53 % of observed access travels), but more as passengers (31 % versus 7 % for active workers). This result has to be linked with the rates of possession of the driver's license and car availability. Overall, students do not walk more than active people (36 % compared to 32 %). Schoolchildren reach a station either by foot or by car (as a passenger of course) but in the same proportion (almost 41 % for each of these two modes). And the other categories of train users are willing to walk more to get to the station (from 48 to 56 % depending on the station). Consequently, for this profile of train users, which relates more to occasional users, the modal shift of the car only reaches 35 % (23 % for car driver).

Relationship between profile of a railway station and modal shares to access it

72If we exclude the fourth cluster with a very small number of stations, we can note that the access travel practices to reach stations do not strongly vary according to the station profile (table 14).

73The car is the first mode used to reach local stations with a modal share around 53 % (clusters 1, 2 and 3). Nearly 37 % of train users drive to the train station and park on site. This proportion is higher for the third cluster in some extent. Going to a station as a passenger of a car concerns 16 % of access travels. Obviously, this value is quite lower for the third cluster.

Table 14: Evolution of modal shares of access modes to station

Walk (%)

Car driver (%)

Car passenger (%)

Other modes (bicycle and public transport) (%)

Cluster 1

Scenario based on actual practices

37

38

17

7

Pro-active scenario

44

35

14

7

Ambitious scenario

44

23

7

25

Cluster 2

Scenario based on actual practices

35

37

15

13

Pro-active scenario

39

34

13

13

Ambitious scenario

39

22

8

30

Cluster 3

Scenario based on actual practices

35

41

13

11

Pro-active scenario

40

38

11

11

Ambitious scenario

40

20

4

36

Cluster 4*

Scenario based on actual practices

47

15

15

24

Pro-active scenario

NS

NS

NS

NS

Ambitious scenario

NS

NS

NS

NS

All stations

Scenario based on actual practices

36

37

16

11

Pro-active scenario

42

34

13

11

Ambitious scenario

42

22

7

29%

* small sample ; NS non significant

74These results represent the current modal shares of travel practices to access a local station. Whatever the profile, the current share of train users driving to stations is relatively high and might justify some actions. In a context of implementing a proactive policy to control car travels, these practices could evolve to the use of more friendly-environmental modes. Indeed, the analysis of travel practices to access stations highlights a large number of short trips made by car. Based on the willingness to walk to a station evaluated according to the distance made by 80 % of walkers (value of 1 km in our sample), we built a first voluntarist scenario of modal shift of car travels (drivers and passengers) who are less than 1 km. It leads to evaluate that potentially 8 % of trips made by drivers and 17 % by car passengers to access a station could be done by foot. In a context of a total modal shift of these short car travels, the modal share of walk would increase from 36 % to 42 % (table 14).

75Then, we have considered a second scenario, which may be more ambitious since it concerns a modal shift of car travels less than 3 km. Travels less than 1km would be done by foot and travels from 1 to 3 km would be done by bicycle. Potentially, considering this range of feeder distance to station, 33 % of car driver trips and 38 % of car passenger trip could be made by bicycle, under the conditions of good cycling and walking accessibility, particularly good pathway quality and a sufficient and secure supply of bicycle parking at stations.

Discussion and conclusion

76This paper presents an assessment of modal shares to access local French train stations. This evaluation provides a methodological approach for local authorities to support accessibility issues of local railway stations, without the need to carry out field surveys. It will help them to objectify forward visions of possible evolution of these current modal shares. It will also help them to measure the real stake of parking offer around stations.

77Everyone agrees that it is necessary to act on railway station accessibility to encourage and increase the train use. The main issue is to understand how to act. There are many works, which try to identify actions levers. For example, the work carried out under the Transit Cooperative Research Program has led to the submission of management recommendations to improve accessibility of stations by all modes (TCRP, 2012). However, before implementing some improvements works, we have to ask the right questions.

78First, a lot of work, in particular this paper, confirm the importance of the feeder distance to station in choosing the train for the daily journeys. We have shown that accessibility to local stations may be globally defined by a radius from 3.5 to 7.9 km around the station. This radius characterizes the influence area, the "connectable'' space to the station (Hasiak et al., 2016a) and also the key territory on which there are challenges to promote the train use. Therefore, it represents the key area to work on railway station accessibility.

79Then, the median value of access travels to station (1.5 km) invites us to consider local stations as stations of proximity. So, given the current feeder distances, the primacy of the car use to reach a railway station should be questioned to the benefit of walking and cycling. As these stations are located in few dense territories, the issue of public transport service may be asked, but it is quite difficult to improve the quality of its service, given the low population density to be served. The challenge of accessibility by the active modes is therefore important. Walking to a station is chosen by 36 % of train users, it is the second access mode after the car, but walking could reach a modal share of 42 % if the short car travels less than 1 km are discouraged!

80Finally, a consensus exists among the scientific community to say that the implementation or extension of car parks does not seem to be the solution to increase the use of railway services (Givoni et al., 2007; Semler et al., 2010). Givoni et al. (2007) suggests that lack of parking facilities near stations may be a reason for a stability of car as an access mode for rail trips. Moreover, Semler et al. (2010) precises that promoting soft modes such as walking, cycling and riding public transport to rail stations might increase ridership without the need to provide additional car parking facilities. This questions public policies of car park offer: should local stakeholders reduce the capacity of car park to promote both train use and friendly access modes? Should they increase the car park offer to enhance the train use? We have to keep in mind that such a policy has significant financial and environmental costs and is also space-consuming. Besides, the surroundings of stations should not be considered as parking spaces. It is therefore necessary to rationalize parking offer by evaluating the real parking demand, taking into account the role of feeder railway stations and the objective of preserving stations environment to allow their territorial valorization (TOD approach). We have appraised that the modal share of the car could decrease from 53 % to 47 % or 29 % depending on the length of the access travel (less than 1 km, less than 3 km). This would lead to a significant reduction in the number of car travels to access stations, representing a saving of up to 40 % in parking places. Encouraging non-motorized modes would have environmental benefits (reduction of noise and pollution) and would contribute to car-free life style (Givoni et al., 2007).

81In addition, there is an urgency to act on behaviors to fight this routine to take his car to get to a station, in particular when the route is small. But, changing routines is a difficult challenge (Buhler, 2012). Yet, considering the challenges of climate, public health and energy transition, which are at the heart of French policies, we have to act on all levers at all scales to promote changes in behaviors and travel practices. There is an important challenge to help people to discover (or rediscover) the benefit of walking or cycling. Therefore, awareness and persuasion programs should be considered by decision makers beyond functional investments to improve accessibility.

82Moreover, improving access to railway stations and encouraging the train use also raise some questions about the capacity of the rail system to absorb an increase of train attendance. Givoni (2007) shows that, in the Netherlands, a 2 % increase in the rail modal share would correspond to a 25 % increase in rail travel demand. Hasiak et al. (2016b) estimate the modal shift from the car to the train around 5 % for regional travels in the region of Picardie, which would mean an increase in train attendance of more than 30 %. This shift will mainly concern railway services with the region of Paris, which are currently at saturation limit.

83Politicians have to work on several levels to enhance the railway transport system and to lead to more virtuous travel behaviors: improving railway stations accessibility, raising awareness of virtuous access travel practices to stations, implementing a good quality rail service (frequency, regularity, comfort). The research community has to work on qualitative approaches to understand the choice of access modes made by train users, especially to understand the brakes to use environment-friendly modes to reach a railway station. It also has to work on tools to impel changes of mobility behaviors. This key challenge requires working on all these fronts in a complementary way.

Haut de page

Bibliographie

Aguilera A., Conti B., Le Néchet F., 2017, "Accompagner la transition vers des mobilités plus durables dans le périurbain", Transports Urbains - mobilités réseaux territoires, No.130, 3-9.

Aguram (Agence d'urbanisme d'agglomérations de Moselle), 2018, "Les haltes ferroviaires sur le périmètre de Metz Métropole", Aguram Publication (https://www.aguram.org), collection REPER' Mobilité.

Audiar (Agence d'urbanisme et de développement intercommunal de l'agglomération rennaise), 2013, "Les pôles d'échanges dans le cœur d'agglomération et le périurbain", Audiar Publication (https://www.audiar.org).

Bavoux J.J., Beaucire F., Chapelon L., Zembri P., 2005, Géographie des transports, Paris, Armand Colin, coll. "U".

Bergman A., Gliebe J., Strathman J., 2011, "Modeling Access Mode Choice for Inter-Suburban Commuter Rail", Journal of Public Transportation, Vol.14, No.4, 23-42.

Brons M., Givoni M., Rietveld P., 2009, "Access to railway stations and its potential in increasing rail use", Transportation Research, Part A, Vol.43, No.2, 136-149.

Buhler T., 2012, "Eléments pour la prise en compte de l’habitude dans les pratiques de déplacements urbains: le cas des résistances aux injonctions aux changements de mode de déplacement sur l’agglomération lyonnaise", thesis report, INSA Lyon.

Carpentier S., 2010, "Modes d'habiter urbains et ruraux: entre continuité et rupture", Articulo – Journal of Urban research, special issue 3 [en ligne], URL: http://articulo.revues.org/1548.

Cerema Nord Picardie, 2018, "Estimation du report modal de la voiture vers l'offre de transport régional – application sur le teritoire du Nord Pas de Calais", report for the Hauts de France region (http://www.observatoire-transports-hauts-de-france.fr).

Cerema Nord Picardie, 2016, "Enquêtes sur les Grands Mobiles Picards: Estimation du report modal de la voiture vers le train", report for the Hauts de France region (http://www.observatoire-transports-hauts-de-france.fr).

Cervero R., Round A., Goldman T., Wu K., 1995, Rail Access Modes and Catchment Areas for the Bart System, working paper, University of California Transportation Center, USA, No. 307.

Cervero R., 2001, "Walk-and-ride: factors influencing pedestrian access to transit", Journal of Public Transportation, Vol.3, No.4, 1-23.

Cervero R., 2002, "Built environments and mode choice: toward a normative framework", Transportation Research Part D, Vol.7, No.4, 265-284.

Cete Nord Picardie, 2012, " Le fonctionnement de pôles d'échanges ferroviaires périurbains pour une accessibilité à la métropole lilloise", report for the Ministry of Transports/DREAL.

Chakour V., Eluru N., 2014, "Analyzing Commuter Train User Behavior: a Decision Framework for Access Mode and Station Choice", Transportation, Vol.41, No.1, 211-228.

Chapuis R., 1995, "L’espace périurbain. Une problématique à travers le cas bourguignon", L’information géographique, No.59, 113-125.

Chen C., Xia J., Smith B., Olaru D., Taplin J., Han R., 2015, "Influence of Parking on Train Station Choice under Uncertainty for park-and-ride Users", Proceedings of the 6th International Conference on Applied Human Factors and Ergonomics (AHFE) and the Affiliated Conferences, Elsevier, Procedia Manufacturing, Vol.3, 5126 – 5133.

Commerçon N., George P., 1999, Villes en transition, Paris, Economica-Anthropos, coll. Géographie.

Corpuz G., Hay A., Merom D., 2005, "Walking for transport and health: trends in Sydney in the last decade", Papers of the 28th Australasian Transport Research Forum, Sydney, Australia, https://www.atrf.info/papers/2005/index.aspx.

Corpuz G., 2007, "Public Transport or Private Vehicle: Factors That Impact on Mode Choice", Papers of the 30th Australasian Transport Research Forum, Sydney, Australia, https://www.atrf.info/papers/2007/index.aspx.

Creemers L., Bellemans T., Janssens D., Wets G., Cools M., 2015, "Analysing Access, Egress, and Main Transport Mode of Public Transit Journeys: Evidence from the Flemish National Household Travel Survey", 94th Annual Meeting of the Transportation Research Board, Washington D.C, USA.

Daniels R., Mulley C., 2011, "Explaining Walking Distance to Public Transport: the Dominance of Public Transport Supply", World Symposium on Transport and Land Use Research, Whistler, Canada.

De Witte A., Hollevoet J., Dobruszkes F., Hubert M., Macharis C., 2013, "Linking modal choice to motility: a comprehensive review", Transportation Research Part A, Part A, Vol.49, 329-341.

Debrezion G., Pels E., Rietveld P., 2009, "Modelling the Joint Access Mode and Railway Station Choice", Transportation Research Part E, Vol.45, 270-283.

Delage A., 2013, "La gare, assurance métropolitaine de la ville post-industrielle – le retournement de valeur dans les projets urbains de quartier de gare à Saint-Etienne Châteaucreux (France) et Liège-Vuillemins (Belgique)", thesis report, Université de Lyon 2 - Lumières.

Department for Transport (UK), 2012, "Public attitudes towards train services: results from the April 2012 Opinions Survey", report.

Dupuy G., 1999, La dépendance automobile: symptômes, analyses, diagnostic, traitements, Paris, Anthropos, coll. villes.

Gallez C., Maulat J., Roy-Baillargeon O., Thébert M., 2015, "Le rôle des outils de coordination urbanisme-transports collectifs dans la fabrique politique urbaine", Flux, Vol.2, No.101-102, 5-15.

Givoni M., Rietveld P., 2007, "The Access Journey to the Railway Station and its Role in the passengers’ satisfaction in the rail travel", Elsevier, Transport Policy, No.14, 357-365.

Habib K.N., Mahmoud M.S., Coleman J., 2013, "Effect of parking charges at transit stations on park-and-ride mode choice: lessons learned from stated preference survey", Journal of the Transportation Research Board, Vol.2351, No.1, 163-170.

Halldorsdottir K., Nielsen O., Prato C., 2017, "Home-End and Activity-End Preferences for Access and Egress from Train Stations in the Copenhagen Region", International Journal of Sustainable Transportation, Vol.11, No.10, 776-786.

Hasiak S., Bodard G., 2016a, "Influence areas of railway stations: how can we explain their geographic forms?", 11th World Congress on Railway Research, Milan, Italy.

Hasiak S., Hasiak F., 2016b, "Understanding rail mobility: are there some links between modal shift to rail services and image of the train?", 11th World Congress on Railway Research, Milan, Italy.

Keijer M.J.N., Rietveld P., 2000, "How do People get to the Railway Station? The Dutch Experience", Transportation Planning and Technology, No.23, 215-235.

Krygsman S., Dijst M., Arentze T., 2004, "Multimodal Public Transport: an Analysis of Travel Time Elements and the Interconnectivity Ratio", Elsevier Transport Policy, No.11, 265-275.

La Paix Puello L., Geurs K., 2014, "Adaptive stated choice experiment for access and egress mode choice to train stations", Proceedings of the World Symposium on Transport and Land Use Research, Delft, the Netherlands.

Le Néchet F., "Consommation d’énergie et mobilité quotidienne selon la configuration des densités dans 34 villes européennes", Cybergeo: European Journal of Geography [En ligne], Systèmes, Modélisation, Géostatistiques, document 529, 18 mai 2011, URL: http://journals.openedition.org/cybergeo/23634.

L'Hostis A., Alexandre E., Appert M., Araud-Ruyant C., Basty M., et al., 2013, "Concevoir la ville à partir des gares", final report, Bahn.Ville 2 project.

Mamoghli M., "Séquences piétonnières et mobilité en transports en commun. Exemple de l'Île-de-France", Revue d’Économie Régionale & Urbaine [en ligne], décembre 2009, URL: https://www.cairn.info/revue-d-economie-regionale-et-urbaine-2009-5-page-995.htm.

Motte-Baumvol B., Belton Chevallier L., Morel-Brochet A., 2012, "Les territoires périurbains entre dépendance automobile et ségrégation socio-spatiale, les ménages modestes fragilisés par les coûts de la mobilité", research report for PUCA (Plan Urbanisme Construction Architecture).

NSW Government, Department of Planning, Transport and Population Data Centre (TPDC), 2006, "Train access and egress modes", Transfigures.

Rak G., Lep M., 2014, "Model of Traffic Access Mode and Railway Station Choice of Suburban Railway System in Slovenia", Transport Problems, Vol.9, No.4, 15-26.

Richer C., Palmier P., 2012, "Mesurer l’accessibilité territoriale par les transports collectifs. Proposition méthodologique appliquée aux pôles d’excellence de Lille Métropole", Cahiers de géographie du Québec, Département de géographie de l’Université Laval, Vol.56, No.158, 427-461.

Roudier E., "Les “Grandes Gares”: des projets adaptés aux villes moyennes?", Métropolitiques, 14 décembre 2015, URL: http://www.metropolitiques.eu/Les-Grandes-Gares-desprojets.html.

Santamaria F., 2012. "Les villes moyennes françaises et leur rôle en matière d'aménagement du territoire: vers de nouvelles perspectives?", Norois, Vol.223, No. 2, 13-30.

Semler C., Hale C., 2010, "Rail station access – an assessment of options", 33rd Australasian Transport Research Forum Conference, Canberra, Australia.

Transit Cooperative Research Program (TCRP): Project B-38 Report 153, 2012, Guidelines for Providing Access to Public Transportation Stations.

Wiel M, 1996, "La mobilité dessine la ville", Urbanisme, No.289, 80-85.

Wiel M., 1999, La transition urbaine: le passage de la ville pédestre à la ville motorisée, Liège, Pierre Mardaga.

Wiel M., 2010, Etalement urbain et mobilité, PREDIT, Paris, La Documentation française.

Zahavi Y., 1973, “The TT-relationship: a unified approach to transportation planning”, Traffic engineering and control, Vol.15, No.10, 205-212.

Haut de page

Table des illustrations

Titre Figure 1: Home to work train travel: a segment of a multimodal transport chain
Crédits Conception S. Hasiak
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-1.png
Fichier image/png, 116k
Titre Figure 2: Classification of French urban areas
Crédits Source: INSEE 2010; conception: S. Hasiak
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-2.png
Fichier image/png, 18k
Titre Figure 3: The French Households Travel Surveys (HTS) used
Crédits Conception Cerema
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-3.png
Fichier image/png, 446k
Titre Figure 4: Our sample of French local train stations
Crédits Conception Cerema
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-4.png
Fichier image/png, 404k
Titre Figure 5: a) frequency of trains at peak hours
Crédits Source: SNCF data 2017
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-5.png
Fichier image/png, 7,4k
Titre Figure 5: b) at off-peak hours
Crédits Source: SNCF data 2017
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-6.png
Fichier image/png, 7,4k
Titre Figure 6: Number of train passengers per day
Crédits Source: SNCF data 2017
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-7.png
Fichier image/png, 13k
Titre Figure 7: Access modes to French regional stations (all travel purposes)
Crédits Sources: open data SNCF 2017 and HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-8.png
Fichier image/png, 196k
Titre Figure 8: Access modes to French local railway stations according to their geographic location
Crédits Source: HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-9.png
Fichier image/png, 7,6k
Titre Figure 9: Modal shares to reach a local station according to access distance (in km)
Crédits Source: HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-10.png
Fichier image/png, 79k
Titre Figure 10: access modes to reach a local station according to gender
Crédits Source: HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-11.png
Fichier image/png, 11k
Titre Figure 11: access modes to local stations according to the age of train users
Crédits Source: HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-12.png
Fichier image/png, 12k
Titre Figure 12: access modes to local stations according to purposes
Crédits Source: HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-13.png
Fichier image/png, 10k
Titre Figure 13: access modes to reach a local station according to departure time
Crédits Source: HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-14.png
Fichier image/png, 12k
Titre Figure 14: The set of indicators characterizing the profile of a train station
Crédits Conception S. Hasiak
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-15.png
Fichier image/png, 17k
Titre Figure 15: Cubic Clustering Criterion (CCC) according to the number of local railway stations clusters (NCL)
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-16.png
Fichier image/png, 9,1k
Titre Figure 16: Typology of station profiles, clustering of stations sample and statistical indicators
Crédits Conception S. Hasiak
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-17.png
Fichier image/png, 262k
Titre Figure 17: Distribution of access distances according to the station profile
Crédits Source: HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-18.png
Fichier image/png, 195k
Titre Figure 18: Distribution of access distances according to access modes
Crédits Source: data from HTS surveys
URL http://journals.openedition.org/cybergeo/docannexe/image/33488/img-19.png
Fichier image/png, 194k
Haut de page

Pour citer cet article

Référence électronique

Sophie Hasiak, « Access Mobility to Local Railway Stations: Current Travel Practices And Forecast », Cybergeo : European Journal of Geography [En ligne], Systèmes, Modélisation, Géostatistiques, document 920, mis en ligne le 21 novembre 2019, consulté le 14 décembre 2019. URL : http://journals.openedition.org/cybergeo/33488 ; DOI : 10.4000/cybergeo.33488

Haut de page

Auteur

Sophie Hasiak

Cerema Hauts de France, France
Project manager, member of the research project-team Esprim
sophie.hasiak@cerema.fr

Haut de page

Droits d’auteur

Licence Creative Commons
La revue Cybergeo est mise à disposition selon les termes de la Licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 3.0 non transposé.

Haut de page