Professional Documents
Culture Documents
a r t i c l e i n f o a b s t r a c t
Article history: Transit-oriented development is being actively promoted as an urban design model for areas around
Received 7 November 2014 transit stations. In addition, planning for accessibility is being promoted, which requires integrating land
Revised 4 March 2015 use with transportation planning, and to match the transportation features with the intensity and diver-
Accepted 28 April 2015
sity of land use of the station areas. Nevertheless, and despite the evident similarities between the two
approaches, an integrated evaluation tool of a station area in terms of its transportation, land use, and
urban design features is missing. In this paper, we bring into the literature on integration of land use
Keywords:
and transport a key feature of the transit-oriented development literature: the urban design features
Transit-oriented development
Node-place model
of the station areas, in particular their pedestrian friendliness. By complementing the node-place model
Walkability with an evaluation of the pedestrian connectivity of station areas of Lisbon, we combine these two per-
Pedestrian connectivity spectives in order to evaluate and classify station areas in three different aspects: land use, transporta-
Lisbon tion, and walkability conditions. Our results show that a balanced node-place is not necessarily a
transit-oriented development, and vice versa, and so a complementary analysis of both is useful to
identify and classify a station area. Therefore, we suggest a typology of station areas based on the three
components, which might be used as a planning tool for the development of the station areas into bal-
anced transit-oriented development areas.
Ó 2015 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jtrangeo.2015.04.009
0966-6923/Ó 2015 Elsevier Ltd. All rights reserved.
D.S. Vale / Journal of Transport Geography 45 (2015) 70–80 71
of stations, either being the original five categories or to propose transit station and associated transportation supply, but it is neces-
other categories based on the land use and transport features of sarily a balanced node-place, in which the development potential
the station areas. conditions have been realized. Indeed, as argued by Calthorpe
Regarding the classification of the overall land use-transport (1993), if a TOD is more than a simple land-use transportation bal-
equilibrium, the initial application of the model to the station areas anced site located around a transit station, the walking environ-
of Amsterdam and Utrecht revealed that the majority of places were ment is a key component of any TOD, allowing the convenient,
located in the balanced area, with a tendency to predominate depen- comfortable and safe walking from and to the station (Jacobson
dent node-places (Bertolini, 1999). In Switzerland, the application of and Forsyth, 2008). Therefore, we argue that by combining the
the model to nearly 1700 station areas revealed that the majority of node-place model with an evaluation of the walkability of each sta-
station areas were located in the ‘unbalanced place’ category, sug- tion area a deeper TOD evaluation can be performed, as not only it
gesting that the majority of station areas are dense and diverse is possible to identify the balance between transport (mainly tran-
(Reusser et al., 2008; Zemp et al., 2011). In Rio de Janeiro, Brazil, sit supply) and land use (as a surrogate of demand), but also the
Gonçalves and Portugal (2008) applied the model to a new train line walking accessibility to the station. We this we intend to bring into
with 14 station areas, and were able to classify them in the five orig- the literature on integration of land use and transport a key feature
inal categories, but the majority were located in the ‘balanced’ and of the TOD and travel behavior literature: the urban design fea-
‘dependent’ category. In Tokyo, Japan, on the contrary, the vast tures of the station areas, in particular their pedestrian friendli-
majority of station areas are located close to the middle line of the ness, i.e. their walkability.
node-place model, revealing a good equilibrium for a large number Supporting the importance of urban design of TODs, research
of station areas, and very few ‘unbalanced nodes’ or ‘unbalanced has been developed to identify the TOD’s urban design features
places’ (Chorus and Bertolini, 2011). In Ostrava, Czech Republic, that have a key importance in the promotion of active travel, liv-
the application of the model revealed several distinct situations, ability and place making. Among other important urban design
but as the city has only 11 station areas, the model is probably very features like variety, complexity and safety, a human scale design
unlikely to provide a clear image of the city (Ivan et al., 2012). Finally, is being pointed as a key feature of a TOD, which is clearly associ-
Kamruzzaman et al. (2014) applied the model to Brisbane, Australia ated with a walkable environment (Jacobson and Forsyth, 2008).
in order to develop a Transit-oriented Development typology. As this Several indicators are found in the literature to measure the walk-
research was applied to the entire city it is not possible to directly ability of a place, from each the walkability index is probably the
compare with the other studies, but through the analysis of the spa- most used (Leslie et al., 2007). In general terms, most walkability
tial distribution of the TOD types it appears that the majority of sta- indicators include three main components: density, land use mix,
tion areas were classified as ‘Potential TODs’, revealing therefore an and connectivity (Frank et al., 2006, 2005). The place-index of
unbalanced situation. the node-place model also measures density and land use mix,
A straightforward method to classify the equilibrium of the and so it is likely that the use of a three-measure walkability index
entire system is to plot in the node-place model diagram the mean would be highly correlated with it. Pedestrian connectivity how-
value for the node- and place-index, and calculate its distance to ever is absent from the model, and so we suggest evaluating it as
the middle line. This will be an important indicator to compare dif- a possible indicator of the walking accessibility to the station.
ferent cities, revealing if the system is located toward the unsus- Operational walkability evaluations of TODs have been devel-
tained node or toward the unsustained place typology. oped for several cases, revealing a great diversity of pedestrian
Unfortunately, the previous applications of the node-place model environments of TODs. Schlossberg and Brown (2004) measured
did not provide the mean value of the indexes, making it impossi- the connectivity of the walking environment, as given by the
ble to compare the cities. However, from the reviewed literature, it pedestrian shed ratio, to classify TODs in Portland, Oregon.
becomes clear that the model is capable of distinguish among sta- Results revealed a considerable variation in pedestrian friendliness
tion areas, revealing different situations of unbalance between land of the sites. Nawrocki et al. (2014) calculated a walkability index
use and transport features of the station areas. for LRT station areas in Japan and USA, and revealed a positive
In terms of typologies of station areas, the literature is not so association with LRT usage in USA but not in Japan. A possible
straightforward in using the five-cases typologies suggested by explanation for the absence of statistical relationship in Japan is
Bertolini (1999). Indeed, based on a cluster analysis, Reusser that the walkability of the majority of LRT stations in Japan is
et al. (2008) suggest a five-cluster solution that do not coincide already successfully established, which is visible in the higher
with Bertolini’s original typologies. Kamruzzaman et al. (2014) walkability score and lower dispersion for the Japanese LRT
suggests four categories of TODs – Residential TODs, Activity Station areas in comparison with the values for USA. Park et al.
Centre TODs, Potential TODs and Non-TODs – although the authors (2014) measured street segments walkability (path walkability)
applied the model to the entire city at the Census Collection to transit stations in Mountain View, California. Four factors were
Districts (CCD) level, and so a direct comparison with the other extracted from the data, revealing a statistically significant relation
studies is not straightforward. Therefore, these differences suggest between walkability to the station and mode choice. Finally, Jiang
that the model is methodologically robust to express the land et al. (2012)’s analysis of the walkable conditions of BRT station
use-transport equilibrium, but the proposed 5-classes solution is areas revealed that the size of the walking distance is influenced
subject to further refinement. by their built environment, suggesting the use of a flexible-size
catchment area in accordance with the walkability features of
2.3. Walking environments of station areas the station area. Overall, all these research findings reveal a great
variety of TODs in terms of pedestrian friendliness, and suggest
One possible enhancement of the node-place model, if used as a that this is an important factor in explaining walking and transit
framework to guide urban and transportation planning to promote usage.
TOD, is to combined it with an evaluation of the urban design fea-
tures of the station areas, as urban design is pointed as a key fea-
ture of TODs (Dittmar and Poticha, 2004; Renne, 2008) and is 3. Methodology
one important dimension to explain travel behavior (Ewing and
Cervero, 2010). This explicitly assumes that a TODs is not simply In this paper we complement the node-place model methodol-
an urban place designed to support and explore the presence of a ogy with an evaluation of the walkability environment of each site.
D.S. Vale / Journal of Transport Geography 45 (2015) 70–80 73
Table 1
Comparison of the indicators used in the reviewed literature to measure node and place index.
Bertolini (1999)a Reusser Gonçalves and Chorus and Zemp et al. Ivan et al. Kamruzzaman Our
et al. Portugal (2008)c Bertolini (2011)e (2012) et al. (2014) methodology
(2008)b (2011)d
Node-index: accessibility of the node
(a) Accessibility by train
Number of directions x x x x y1
served
Daily frequency of services x x x x x y2
Number of stations within x
45 min of travel
Number of stations within x x x y3
20 min of travel
Type of train connections x
(b) Accessibility by bus, tram and underground
Number of directions x x x x y4
served
Daily frequency of services x x n.a. x x y5
Daily frequency of x
suburban buses
(c) Accessibility by car
Distance from the closest x x n.a. y6
motorway access
Car parking capacity x n.a. n.a. x y7
(d) Accessibility by bicycle
Number of free-standing x
bicycle paths
Bike path length within x n.a.
2 km
Bicycle parking capacity x n.a. n.a.
(d) Other measures
Public Transport x
Accessibility Level
(PTAL)
Case study Amsterdam and Switzerland Rio de Janeiro, Tokyo, Japan Switzerland Ostrava, Brisbane, Lisbon,
Utrecht, Brasil Czech Rep. Australia Portugal
Netherlands
D.S. Vale / Journal of Transport Geography 45 (2015) 70–80 75
Table 1 (continued)
Bertolini (1999)a Reusser Gonçalves and Chorus and Zemp et al. Ivan et al. Kamruzzaman Our
et al. Portugal (2008)c Bertolini (2011)e (2012) et al. (2014) methodology
(2008)b (2011)d
Number of cases (station 31 1684 14 99 1700 11 1734f 83
areas)
Buffer size 700 m 700 m 700 m 700 m 700 m 700 m 800 m 700 m
Total indicators 15 11 8 9 6 11 6 13
a
Zweedijk (1997) and Serlie (1998) operationalized the indicators.
b
All indicators were log-transformed except bike path length and degree of functional mix.
c
Gonçalves and Portugal (2008) followed Reusser et al. (2008), but due to data limitations had to exclude some indicators.
d
The indicators ‘number of train connections’ and ‘workforce’ were log-transformed.
e
A total of 10 indicators were used to compute a density-use model, but only 6 were used to compute the node-place model.
f
This study was applied at the Census Collection District (CCD) level, for the entire city.
Table 2
Indicators used to measure node- and place-index.
values are small, revealing that the pedshed ratios show a great useful to distinguish among ‘balanced’ places, once in this group
variability, especially for station areas with intermediate node is possible to identify very diverse situations, in which the node
and place indexes. and place indexes are not related with the pedshed ratio.
We have developed a cluster analysis in order to identify the 5. Discussion and conclusions
usefulness of including an evaluation of the walkability features
of station areas in conjunction of the node-place model. Several From a strictly land use and transportation balance point of
hierarchical cluster analyses were developed, using squared view, the node-place model adequately identifies and classifies
Euclidean distance as dissimilarity measure, and changing the clus- the station areas of LMA. It clearly suggests that, for the vast major-
ter method: minimum distance, maximum distance, ity of station areas, the transportation supply is not enough to
between-groups distance and Ward’s method. In all methods seven match the potential demand created by the existent land uses
clusters was the optimum solution, in accordance with a visual around the stations (mean node index = 0.456; mean place
analysis of the scree plot and the R2 criterion. A subsequent index = 0.543). Indeed, in LMA, and especially in the North Bank,
non-hierarchical k-means cluster analysis was performed for seven the suburban growth was mainly structured by the train, which
clusters, explaining 75.9% of the total variance (R2 = 0.759). has lead to the creation of urban ‘‘areolas’’ located around the sta-
In accordance to the values of the cluster centers and tions. The car-driven suburbanization started only after the begin-
F-statistics (Table 3), the place index is the variable that more ning of the 1990s following significant public investments in the
allows to identify the clusters (F = 47.4), followed by PedShed road network (Barata Salgueiro, 2001). However, the node-place
(F = 42.0). The node index is the variable that less identifies the model is revealing the actual land use and transport conditions
clusters (F = 32.5). The seven clusters show some consistency in favoring car travel, once the public transportation supply is clearly
the three variables, and they are placed in a certain order in accor- insufficient and does not take advantage of the urban development
dance with the node and place chart (Fig. 7). Indeed, ‘stressed’ concentrated in the sites around the stations. The results can there-
places show above average pedshed ratios (cluster 4), and ‘depen- fore be considered adequate to classify the node-place equilibrium
dent’ places show very small pedshed ratios (cluster 3). of the urban system, and accurately identify the unbalanced
Nevertheless, the pedshed ratio analysis of the node-places is situations.
76 D.S. Vale / Journal of Transport Geography 45 (2015) 70–80
Fig. 3. Examples of calculated pedestrian shed ratios for station areas of Lisbon Metropolitan Area.
However, it appears that the node-place model is somehow in terms of urban form, land use and activities, and also travel con-
insufficient to distinguish among the ‘balanced’ areas, as this bal- ditions. Indeed, our results show that a fine-grain analysis of the
ance can represent a variety of distinct built environment realities, pedestrian accessibility of the node-places is an important addition
D.S. Vale / Journal of Transport Geography 45 (2015) 70–80 77
Fig. 4. Node-place values of the station areas of Lisbon Metropolitan Area, 2013.
Fig. 5. Node index, place index and squared differenced of the station areas of LMA, 2013.
to the evaluation of the station areas of a metropolitan area. This is Our cluster analysis results suggest a six categories typology of
particularly useful to distinguish among the balanced node-places, station areas based on node index, place index and pedshed ratio.
and therefore to provide a framework that can more easily relate to The first group comprises ‘pedestrian friendliness balanced
Transit Oriented Development, by explicitly evaluating the physi- node-places’ (cluster 4), in which not only there is a balance
cal characteristics of station areas, i.e. the design dimension of between node and place, but also the walking accessibility is high.
the seven Ds framework, and not only physical proximity. We designated this group as ‘Urban TODs’, as they have high
78 D.S. Vale / Journal of Transport Geography 45 (2015) 70–80
Table 3
Final cluster centers and F statistic for each dimension.
1.00
‘‘Balanced TADs’’, corresponding to places revealing node-place
Cluster
Number balance, but with poor pedestrian accessibility (cluster 6). There
1 is physical proximity between the station and the destinations,
2
.80 3 but walking conditions are poor. We identify a third group as
4
5 ‘Suburban TODs’, and is constituted by areas with very high ped-
6 shed ratios and average node and place index (cluster 2). They
7
.60
are places with good walking accessibility to the station and pre-
sent a balance of node and place indexes, but with values lower
Node
This typology can be used as a strategic framework for urban and Nevertheless, we consider that the inclusion of the pedestrian shed
transportation planning of the station areas. In general terms, the ratio will in some extent compensate for this lack of detail, keeping
typology suggests that ‘Urban TODs’ and ‘Suburban TODs’ are the node-place model fast and simple to compute. Fourth, our TOD
already balanced node-places with good walking accessibility con- typology should be used with care, as it reflects only the situation
ditions, in which the transportation supply should not decrease. of LMA and an external validation is needed before any final con-
Likewise if new major activities or urban facilities are to be located clusion. A further validation of typologies based on performance
in these places, a transportation impact assessment is suggested. indicators such as modal share data is needed, as well as the appli-
For ‘Balanced TADs’, an improvement of the walking conditions is cation of this three-component methodology to other cases. The
suggested. ‘Undersupplied transit TODs’ need an increase in transit use of experts can also be a useful validation tool for the proposed
supply, and therefore the main intervention should come from typology. Finally, we recognize that the use of the pedshed ratio as
transportation planning. For these two categories, the impact of an indicator of the walkability of an urban area does not fully
urban and/or transportation measures on transit patronage and incorporates the diverse built environment features that can influ-
active travel can be significant, as these places are already balanced ence walking (Ewing and Handy, 2009; Lee and Moudon, 2004,
node-places. ‘Unbalanced TODs’ should be given special attention 2006; Talen and Koschinsky, 2013). Therefore, other variables
from urban planning, as new travel generating activities could be should be tested to measure the pedestrian friendliness of a
guided to these locations, realizing the available transit node-place balanced station area, which might constitute a real
over-supply. Nevertheless, new transportation supply and/or transit-oriented development site.
improvement of walking conditions should be required as the result
of these new activities. Finally, ‘Car dependent node-places’ require Acknowledgments
an improvement in transportation supply, land use mixing and also
walking conditions to become transit-oriented development. A clear We would like to acknowledge the financial support given to
articulation and integration of land use and transportation planning CIAUD by the Portuguese Funding Agency for Science, Research
is particularly relevant if these places are to be developed. and Technology (project reference PEst-OE/EAT/UI4008/2013),
Nevertheless, especially in a suburban context in which the train is and a special recognition to Miguel Saraiva and Mauro Pereira for
a key transportation mode for commuting, they constitute better their help in collecting the transit supply data.
urban development places than other places located farther away
from stations, in which travel is strictly dependent on the car and References
mass transit will require major financial investments.
Barata Salgueiro, T., 2001. Lisboa, Periferia e Centralidades [Lisbon, Periphery and
Centralities]. Celta Editora, Oeiras.
5.1. Limitations and further developments Barton, H., 1998. Eco-neighbourhoods: a review of projects. Local Environ. 3, 159–
177.
Despite the usefulness of the addition of the walking environ- Bertolini, L., 1996. Nodes and places: complexities of railway station
redevelopment. Eur. Plan. Stud. 4, 331–345.
ment evaluation to the node-place model as an urban planning Bertolini, L., 1999. Spatial development patterns and public transport: the
tool, we recognize some limitations in our results. First, the model application of an analytical model in the Netherlands. Plan. Pract. Res. 14,
seems very sensitive to the normality of the distribution of the 199–210.
Bertolini, L., 2005. Cities and transport: exploring the need for new planning
variables used to measure node and place indexes. Nevertheless, approaches. In: Albrechts, L., Mandelbaum, S.J. (Eds.), The Network Society: A
we argue that, if necessary, the variables should be New Context for Planning. Routledge, London and New York, pp. 67–80.
log-transformed to increase their normality, because they are Bertolini, L., 2008. Station areas as nodes and places in urban networks: an
analytical tool and alternative development strategies. In: Bruinsma, F., Pels, E.,
measured at different scales, and all have the same weight in the Priemus, H., Rietveld, P., Van Wee, B. (Eds.), Railway Development: Impacts on
calculation of the indexes. However, this option, also done by other Urban Dynamics. Physica-Verlag Heidelberg, Leipzig, pp. 35–57.
researchers (Chorus and Bertolini, 2011; Reusser et al., 2008), Calthorpe, P., 1993. The Next American Metropolis: Ecology, Community and the
American Dream. Princeton Architectural Press, New York.
seems to increase the number of places being classified as ‘bal- Cervero, R., Murphy, S., Ferrell, C., Goguts, N., Tsai, Y.-H., 2004. Transit-Oriented
anced’. Therefore, in order to precisely understand the impact of Development in the United States: Experiences, Challenges and Prospects –
the data transformation on the classification of station areas, a sen- TCRP Report 102. Transportation Research Board, Washington, DC.
Chorus, P., Bertolini, L., 2011. An application of the node place model to explore the
sitivity analysis of the results of the node-place model applied to
spatial development dynamics of station areas in Tokyo. J. Transp. Land Use 4,
distinct metropolitan areas is suggested as future research, which 45–58.
can bring some external validity of these findings. Second, as Dittmar, H., Poticha, S., 2004. Defining transit-oriented development: the new
shown in our research and in other previous applications of the regional building block. In: Dittmar, H., Ohland, G. (Eds.), The New Transit
Town: Best Practices in Transit-Oriented Development, pp. 19–40.
node-place model, it seems that a correlation between node and Ewing, R., Cervero, R., 2010. Travel and the built environment: a meta-analysis. J.
place indexes is somehow present, leading to a vast majority of sta- Am. Plan. Assoc. 76, 265–294.
tion areas being located in the ‘balanced’ category and/or along the Ewing, R., Handy, S., 2009. Measuring the unmeasurable: urban design qualities
related to walkability. J. Urb. Des. 14, 65–84.
middle line. Although theoretically the five typologies might exist, Frank, L.D., Schmid, T.L., Sallis, J.F., Chapman, J., Saelens, B.E., 2005. Linking
the previous results suggest that probably unbalanced nodes are objectively measured physical activity with objectively measured urban form:
very hard to identify in practice, suggesting that the influence of findings from SMARTRAQ. Am. J. Prev. Med. 28, 117–125.
Frank, L.D., Sallis, J.F., Conway, T.L., Chapman, J.E., Saelens, B.E., Bachman, W., 2006.
transportation supply is higher than the influence of the land use Many pathways from land use to health: associations between neighborhood
demand. An alternate explanation is that the balance between walkability and active transportation, body mass index, and air quality. J. Am.
transport (node) and land use (place) might not be given by the Plan. Assoc. 72, 75–87.
Gonçalves, J.A.M., Portugal, L.d.S., 2008. Classificando estações metro-ferroviárias
middle line, but instead by a concave curve representing the best como pólo promotor do desenvolvimento socioeconômico, 4° Concurso de
fit regression line (Reusser et al., 2008). This solution would how- Monografia CBTU 2008 – A Cidade nos Trilhos. Companhia Brasileira de Trens
ever decrease the usefulness of the node-place model to compare Urbanos, Rio de Janeiro.
Hale, C., 2012. TOD versus TAD: the great debate resolved. . .(?). Plan. Pract. Res., 1–
different territories, and so we suggest, as an alternative simple
16
solution, to include and reveal the mean node index and place Ivan, I., Boruta, T., Horák, J., 2012. Evaluation of railway surrounding areas: the case
index values as two essential features of the urban system being of Ostrava city. In: Longhurst, J.W.S., Brebbia, C.A. (Eds.), Urban Transport XVIII
analyzed. Third, the model uses 700 m straight-line catchment – Urban Transport and the Environment in the 21st Century. WITPress, pp. 141–
152.
areas for each station to measure the place index, and network dis- Jacobson, J., Forsyth, A., 2008. Seven American TODs: good practices for urban
tance catchment areas could lead to a reduction of the place index. design in transit-oriented development projects. J. Transp. Land Use 1, 51–88.
80 D.S. Vale / Journal of Transport Geography 45 (2015) 70–80
Jiang, Y., Christopher Zegras, P., Mehndiratta, S., 2012. Walk the line: station Renne, J.L., 2008. From transit-adjacent to transit-oriented development. Local
context, corridor type and bus rapid transit walk access in Jinan, China. J. Environ. 14, 1–15.
Transp. Geogr. 20, 1–14. Renne, J.L., Bartholomew, K., Wontor, P., 2011. Transit-oriented and joint
Kamruzzaman, M., Baker, D., Washington, S., Turrel, G., 2014. Advance transit development: case studies and legal issues. Transit Cooperative Research
oriented development typology: case study in Brisbane, Australia. J. Transp. Program, Legal Research Digest 36. National Research Council, Washington, D.C.
Geogr. 34, 54–70. Reusser, D.E., Loukopoulos, P., Stauffacher, M., Scholz, R.W., 2008. Classifying
Landis, J., Cervero, R., Hall, P., 1991. Transit joint development in the USA: an railway stations for sustainable transitions – balancing node and place
inventory and policy assessment. Environ. Plan C: Govern. Pol. 9, 431–452. functions. J. Transp. Geogr. 16, 191–202.
Lee, C., Moudon, A.V., 2004. Physical activity and environment research in the Schlossberg, M., Brown, N., 2004. Comparing transit-oriented development sites by
health field: implications for urban and transportation planning practice and walkability indicators. Transp. Res. Rec. 1887, 34–42.
research. J. Plan. Lit. 19, 147–181. Serlie, Z., 1998. Stationslocaties in vergelijkend perspectief. Master Thesis,
Lee, C., Moudon, A.V., 2006. The 3Ds+R: quantifying land use and urban form Universiteit Utrecht.
correlates of walking. Transport. Res. D: Transp. Environ. 11, 204–215. Talen, E., Koschinsky, J., 2013. The walkable neighborhood: a literature review. Int. J.
Leslie, E., Coffee, N., Frank, L.D., Owen, N., Bauman, A., Hugo, G., 2007. Walkability of Sustain. Land Use Urban Plan. 1, 42–63.
local communities: using geographic information systems to objectively assess Wegener, M., Fürst, F., 1999. Land-use Transport Interaction: State of the Art.
relevant environmental attributes. Health Place 13, 111–122. Institut für Raumplanung, Universitat Dortmund, Dortmund.
Nawrocki, J., Nakagawa, D., Matsunaka, R., Oba, T., 2014. Measuring walkability and Zemp, S., Stauffacher, M., Lang, D.J., Scholz, R.W., 2011. Classifying railway stations
its effects on Light Rail usage: a comparative study of the USA and Japan. In: for strategic transport and land use planning: context matters! J. Transp. Geogr.
Brebbia, C.A. (Ed.), Urban Transport XX. Wessex Institute of Technology, UK, pp. 19, 670–679.
305–316. Zweedijk, A., 1997. Knoop of Plaats? Naar een operationalisering van het begrip
Park, S., Choi, K., Lee, J.S., 2014. To walk or not to walk: testing the effect of path stationslocatie. Master thesis. Universiteit Utrecht.
walkability on transit users’ access mode choices to the station. Int. J. Sustain.
Transport.