Journal of Transport Geography 35 (2014) 111

Contents lists available at ScienceDirect

Journal of Transport Geography

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

Accessibility modelling: predicting the impact of planned transport

infrastructure on accessibility patterns in Edinburgh, UK
Saleem Karou , Angela Hull 1
School of the Built Environment, William Arrol Building, Heriot-Watt University, Riccarton, Edinburgh EH14 4AS, United Kingdom

a r t i c l e

i n f o

Keywords:
Accessibility
Transport planning
Spatial equity
Edinburgh Tram
Geographic Information System (GIS)

a b s t r a c t
The achievement of good spatial accessibility and equity in the distribution of urban services is one of the
supreme goals for urban planners. With Scottish Government backing, the City of Edinburgh Council
(CEC) has started to construct a tram network to cater for the future needs of Scotlands capital city by
providing an integrated transport solution using trams and buses. Spatial Network Analysis of Public
Transport Accessibility (SNAPTA) which is a GIS-based accessibility model has been developed to measure the accessibility by public transport to different urban services and activities. The model responds
to several limitations in other existing accessibility models in planning practice. It offers an alternative
and practical tool to help planners and decision makers in examining the strengths and weaknesses of
land use transport integration. SNAPTA has been applied to a pilot study in Edinburgh city to identify
the contribution of the infrastructure improvements of the tram system and Edinburgh South Suburban
Railway (ESSR) to improved accessibility by public transport to six types of activity opportunities. This
paper outlines the concept and methodology of the SNAPTA model, and presents the ndings related
to this pilot study with a focus on changes in potential accessibility to jobs between four different public
transport network scenarios. The accessibility values so obtained help to identify the gaps in the coverage
of the public transport network and the efciency in the spatial distribution of urban services and activities. The ndings focus on whether the planned transport infrastructures for Edinburgh will lead to better accessibility and reduced inequity (in terms of accessibility) across the city.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
Being able to reach the spatial opportunities in the city-region
where you live without too much hassle is considered as one of
the dimensions of quality of life in empirical studies of life quality
(Bowling and Windsor, 2001; Leitmann, 1999; Roseland, 1997).
The ability to access necessary services is a function of the range
of transportation choices available and their travel time, safety,
cost, and convenience as well as the internal structure of settlements and the spatial distribution of opportunities (Banister and
Hickman, 2007; Forward, 2003). The efcient connection of the
distributed infrastructure of services and facilities with the infrastructure for movement across city regions is a pressing issue for
urban managers. The changing intensity of development at locations in the city-region affects travel demand and the performance
of the transport system whilst city scale transportation investment
alters the accessibility of different parts of the city-region (Banister
and Hickman, 2007; Chapin and Kaiser, 1979; Himanen et al.,
Corresponding author. Tel.: +44 (0) 77 3558 1551.
E-mail addresses: [email protected],
[email protected] (A. Hull).
1
Tel.: +44 (0) 131 451 4407.

[email protected]

(S.

0966-6923/$ - see front matter 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jtrangeo.2014.01.002

Karou),

2005; Holl, 2006; NICHES, 2007; OSullivan, 1980; Priemus et al.,

2001; Sultana, 2006). The dialectical relationship between transport services and spatial opportunities affect both accessibility
and spatial equity, another concept closely linked to quality of life.
The role that public transport plays in connecting communities
and neighbourhoods and the impact of transport investment on
those same communities is acknowledged in local transport policies that seek, for example, To improve the transport choices
households have available to reach a range of services or To promote accessibility to everyday facilities for all, especially for those
without a car (Hull and Karou, 2011). The spatial growth of urban
areas and the decentralization of employment and facilities have
made it harder for people without access to a car to make the daily
commute and to take advantage of distributed retail and leisure
opportunities.
In this respect, there has been a growth of interest in the concept of accessibility over the last decades, with many accessibility
studies published in the academic press discussing how to measure
accessibility and the contribution such decision support tools
might have. Recently, the development of accessibility models
has used a multitude of approaches to inform land use and transport decision-making (Karou and Hull, 2012). Therefore, translating the concept of accessibility into a practical planning tool

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

stems from the need for powerful techniques to help planners and
decision makers deal with urban and transport management and
provide better evaluation of the impacts of different schemes (or
combinations of schemes) advanced by transport and land-use
policies.
This paper focuses on accessibility addressing issues of spatial
equity and transport disadvantage through two objectives. The
rst objective is to develop an accessibility model the Spatial
Network Analysis of Public Transport Accessibility (SNAPTA)
which has responded to the need for academic research models
to be more practical and useful models for the world of planning
practice. The second objective is to test the model through empirical study in the city of Edinburgh based on ex ante evaluation of
the new tram system and Edinburgh South Suburban Railway
(ESSR) to compare between the accessibility impacts of different
scenarios of the completion of these infrastructures.
The paper is organised in six sections. The introduction has
identied transport accessibility as a key dimension of quality of
life and a priority for sustainable urban management. This
acknowledges the interaction between land use and intensity, individual travel behaviour and transport provision. The next section
introduces the case study of Edinburgh. Section 3 discusses the
rationale for the construction of the tram system and re-opening
of ESSR. In Section 4, the conceptual framework and theoretical
underpinning of the SNAPA model is presented. Section 5 focuses
on the methodology of SNAPTA application to Edinburghs network
while the last two sections outline the ndings and further developments in SNAPTA.
2. Case study of Edinburgh
The city of Edinburgh is situated in the central urban belt of
Scotland with an overall density of 37.65 persons per hectare
(2001 census). The policies in the land use plan and Edinburghs
geographical location (bordered by the Firth of Forth on two sides)
have contained urban sprawl, through the imposition of a green
belt around the urban area and the encouragement of development
on browneld sites.

Edinburghs population is projected to grow by over 59,000 between 2010 and 2030 (CEC, 2010). As Edinburghs population
grows, the demand for travel will increase. Population growth in
the city region will also impact on levels of commuting into the
city. Moreover, during the next 20 years, Edinburghs economy is
forecast to play a big part in Scottish economic growth (CEC,
2010). The city is currently commencing a huge phase of redevelopment. Edinburgh Waterfront is set to provide an additional
25,800 new residential units and nearly 350,000 m of new ofce,
retail and other commercial developments between 2006 and
2020. Signicant new development is also predicted to be progressively built by 2020 in West Edinburgh with some 250,000 m of
new ofce space and over 200,000 m of other commercial space
(TIE, 2006). Fig. 1 shows the location of housing and ofce
developments programmed for completion between 2006 and
2015 based on outstanding consents and local plan allocations
(CEC, 2008).
Continuing economic success has however created a number of
challenges. With a substantial population increase expected and
The number of jobs. . .. . ..now expected to increase by 15% between 2000 and 2015 (CEC, 2007, p. 14) as well as the forecast rise
in household car ownership by 30% from 2000 and 2016 causing
twice as much time to be lost due to congestion over the same
period (TIE, 2004, p. 2), the maintenance of connectivity and accessibility is one such challenge (Hull and Karou, 2011). The Transport
2030 Vision argues that, by 2030, without action, the demand for
travel from/to the city by private car will far exceed the current
capacity (CEC, 2010).
The City of Edinburgh Council has dened a series of actions
including the implementation of new public transport infrastructures such as the tram system and ESSR to boost the transport
system and improve accessibility in the Councils area. The expectation is to cut demand for road travel and to serve the new growth
areas while they develop by delivering a reliable and safe public
transport service and, consequently, by improving their accessibility. The Public and Accessible Transport Action Plan (PATAP) 2013
2020 suggests that the target is to increase public transports share
of all their journeys by 2015 by 1.3%, and by 2020 by 2.3%

Fig. 1. Location of housing and ofce developments programmed for completion by 2015. Source: City of Edinburgh Council planning records (2008).

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

compared to the average attained between 20072008 and 2009

2010 (19.1%) (CEC, 2013a, p. 25).
The Scottish Government perceives high accessibility as essential to economic growth and competitiveness through providing
access to markets and enhancing the attractiveness of cities as focal business locations and tourism (Scottish Executive, 2004, p.
18). In the National Transport Plan, accessibility is linked to
improving journey times and connections and to the quality and
affordability of public transport choices (Scottish Executive,
2006a, p. 2). Accessibility is translated into the Edinburgh Local
Transport Strategy as whether or not people can get to services
and activities at a reasonable cost, in reasonable time and with reasonable ease (CEC, 2007, p. 82).
A study carried out by MVA Consultancy (2008) in association
with SEStran (South East Scotlands Regional Transport Partnership) used Accession model to look at accessibility to the key hospitals and employment sites in the region. The study focused on
the calculation of travel times using walking, cycling, car and public transport before the construction of Edinburgh Tram. Two other
previous studies of accessibility, before the development of the
tram and ESSR, examined the transport and land use effects of major new land use developments in the Edinburgh city-region. Derek
Halden Consultancy (2002) examined how accessibility to jobs
would change if a proportion (20%) of future development (development not already committed) was allocated according to different spatial strategies (e.g. green belt development; development of
new settlements, etc.). David Simmonds Consultancy used a bespoke version of TELMoS to predict the impact of two major new
strategic headquarters developments to the west of Edinburgh beyond the city bypass close to the airport (Bramley et al., 2011).
These two studies have identied two highly policy-relevant considerations for CEC. Firstly, the public transport underperformance
in the north western zone of the city towards the city bypass which
particularly affects zones of affordable housing (Halden, 2002).
Secondly, that the development in one area outside the city bypass
has an impact, in terms of congestion, pollution and trafc levels
throughout a much wider geographical area.

3. The rationale for the Edinburgh Tram and South Suburban

Railway
The Edinburgh Tram was rst mooted in the 1990s and received
parliamentary assent in March 2006. The Edinburgh Local Transport Strategy 20072012 denes the tram scheme as the key project coming to Edinburghs transport network, emphasising that
the Council is committed to implementing the project to strengthen the city public transport system. The Local Strategy argues that
for the Edinburgh Tram to be successful and attract people, it will
require full integration with existing bus services (i.e. through
common ticketing, interchange points and timetabling), and with
the fabric of the city (CEC, 2007). It is intended that Tram ticketing
will be integrated with Lothian Buses covering day and season tickets. However, full ticket integration, e.g. where any bus service
feeding into the tram provides a simple through ticket even for single journeys, cannot currently be delivered due to legislative
restriction (CEC, 2013b).
The tram, which is being delivered by Transport Initiatives
Edinburgh (TIE) a company formed by CEC, is currently under
construction with the completion date having been deferred on
numerous occasions due to legal action concerning the nancial
costs, disturbance and upheaval costs.
The original 2001 proposal for Edinburgh Trams envisaged
three lines across the city; the rst being a circular route running
around the northern suburbs, with the other two forming radial
lines running out to Newbridge in the west and to Newcraighall

in the south east respectively (CEC, 2006). All lines would run
through the city centre. After Line Three was shelved, Lines One
and Two were combined and split into three phases, with Phase
1 being further divided into Phase 1a and 1b (see Fig. 1), as follows:

Phase
Phase
Phase
Phase

1a; Newhaven to Edinburgh Airport.

1b; Haymarket to Granton Square.
2; Newhaven to Granton.
3; Edinburgh Airport to Newbridge.

As a result of the suspension of work on Line Three due to lack

of Scottish Parliamentary approval and later on Phases 1b, 2 and 3
due to lack of funding (CEC, 2011), in September 2011 only the
construction of part of Phase 1a from the Airport to central Edinburgh was started. However, the intention is to secure funding
for the additional lines (CEC, 2013b).
West Edinburgh from the Gyle shopping centre to Newbridge
has been identied by the Scottish Government as a national
growth point. Tram Phase 1a at 18.5 km in length is, therefore, seen
as vital to linking the 56 hectare development site at Leith through
West Edinburgh growth point to the airport and in responding to
the expected growth in travel demand (TIE, 2007, p. 41). The Business Case for the tram argues that the likely success of the development between Granton and Leith (Fig. 2), and therefore the
CEC strategy, will be strongly affected by the provision of a reliable,
sustainable public transport network, of which the tram plays an
essential part (TIE, 2007, p. 41). The Business Case adds that in
the absence of the tram Phase 1a and Phase 1b, the new proposed
development in North Edinburgh may be diverted to less sustainable locations with less potential for successful transport integration (TIE, 2007, p. 41).
The Edinburgh South Suburban Railway (ESSR) is an existing
double track railway line passing through the suburbs to the south
of the city centre which is used by freight trafc crossing the city.
The feasibility of reopening of the ESSR to passenger services,
which were withdrawn in 1962, has been considered in a recent
study for CEC by Atkins (2004).
Journey to work data shows that the corridor around south central Edinburgh in which the ESSR runs has high levels of public
transport use, particularly to the city centre, but also for many
peripheral journeys further aeld (CEC, 2008). A number of objectives have been dened by CEC (2008) and Transform Scotland
(2007) for the ESSR project to contribute to the wider strategy of
the region and city. These include transforming cross-city links;
improving accessibility to designated employment growth areas;
provide an important feeder to Waverley Station and the programmed new bus/tram/train interchange at Haymarket; making
a signicant shift in peak period journey-to-work trips from the
car to public transport; enhancing the connections between the
areas served by ESSR and other public transport modes (i.e. Edinburgh Tram, the national rail network and bus services); ensuring
access for all potential users to any new services or infrastructure;
and minimising the environmental impacts of travel in the corridor
of the railway (CEC, 2008; Transform Scotland, 2007).
The Atkins study in 2004 concluded that the most feasible option in the short- to medium-term would be to extend the existing
North Berwick Waverley/Haymarket services to Niddrie (see
Fig. 3) (Atkins Transport Planning, 2004). However, the Atkins
2004 report argued that the construction of Line Three of Edinburghs proposed tram system to the south east of the city would
clearly reduce demand levels and signicantly erode the case for
the scheme since it would compete with the locations of planned
stations on the ESSR (Atkins Transport Planning, 2004).
For CEC, however, the extent to which the tram and ESSR will
attract current and future car drivers to public transport is critical.
Also pertinent is how they will contribute to improved accessibility

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

Fig. 2. Edinburgh Tram network. Source: http://www.edinburghtrams.com.

Fig. 3. Edinburgh South Suburban Railway (ESSR) re-opening proposal. Source: TRANS form Scotland (2007, p. 2).

and affect the relationships between local travel and activity

choices. These latter issues are the subject of this research.
4. Conceptual framework and theoretical underpinning
Although many accessibility models have been recently developed and tested in scientic research (e.g. Gutirrez and Gmez,
1999; Geurs and van Eck, 2001; Halden, 2002; Yigitcanlar et al.,
2007; Curtis and Scheurer, 2010), the usability of accessibility
models in planning practice is a much less-developed area of study.
Many models are restricted to academic studies due to the complexity of their theoretical underpinnings which leads to a level

of detail and complication that makes their output difcult for policy makers and practitioners to understand and interpret. Other
models have been abandoned due to several failures or limitations
related to operational and methodological issues. For example,
some accessibility models are based on an inadequate theoretical
basis or methodology by relying on very simple or inaccurate
accessibility measures which either are not sensitive to changes
in both the transport system and the land-use system, or fail to reect actual travel behaviour.
Karou and Hull (2012) reviewed a number of current accessibility
models, including: PTAL (London Borough of Hammersmith and Fulham), WALC (Transport Studies Group University of Westminster),

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

PTAM (West Yorkshire Passenger Transport Executive), CAPITAL

(Transport for London), TRANSAM (Brown & Root), SONATA (SDG),
Accession (MVA and Citilabs), SNAMUTS (Carey Curtis and Jan Scheurer), GenMod (Transportation Planning Department of Amsterdam), TMfS (MVA and David Simmonds Consultancy), ACCALC
(Derek Halden Consultancy), LUPTAI (Tan Yigitcanlar and colleagues) and Space Syntax (University College London) and identied some limitations. Some of these models are inexible and non
user friendly in such a way which requires high modelling skills or
a lot of time to operate and input or update the data. Other models
require an external function to be integrated into the GIS environment which might be very expensive and needs a high level of expertise in operating the software. Being restricted to only one transport
mode is another common limitation. In addition, most of the existing
models have failed, somehow, to consider a number of issues in connection with how people perceive accessibility, including: measurement of the actual walk access time (or distance) when connecting
with public transport or the private car; inuence of physical features (e.g. slope); inuence of trafc congestion; interchange option
of public transport journeys between different modes or operators;
inuence of travel at specic times of day (i.e. peak time or off-peak
time) and on specic days of the week (i.e. during weekday or the
weekend); inuence of the signicance of urban activities; and the
declining attractiveness of activities with increasing spatial separation (i.e. travel time or distance).
Several studies have investigated how to choose an appropriate
accessibility measure or model and evaluate the usefulness of its
application in planning practice (see Morris et al., 1979; Koenig,
1980; Cervero et al., 1995; Handy and Niemeier, 1997; Reneland,
1998; Halden et al., 2000; Ross, 2000; Geurs and van Eck, 2001;
Geurs and van Wee, 2004; Keller et al., 2012). Although it is clear
that there is no one best method for assessing accessibility, reviewing the literature revealed a number of issues that characterise the
usefulness of an accessibility model for a particular situation in
planning practice, as follows:
Robustness of theoretical basis, providing an adequate representation of accessibility aspects, with a rational method
of calculation.
Sufcient level of data disaggregation.
Not complex, simply operated, oriented towards clear
objectives.
Easily interpreted, understood and communicated with
planners, researchers and policy makers.
SNAPTA is a GIS-based accessibility model which denes accessibility as whether or not people can get to services and activities
at a reasonable cost, in reasonable time and with reasonable ease.
It offers better usability, covering aspects of accessibility adequately without making it very difcult to operate, interpret and,
consequently, apply in practice. However, the model does not
claim to provide the complete picture of actual travel behaviour
and transport accessibility. It attempts to achieve a balance between the ease of interpretation and operationalisation and the
complexity of the theoretical basis and data disaggregation. The
performance of SNAPTA has been measured against the four
above-stated criteria for creating a useful accessibility model in
planning practice.
SNAPTA is intended to assist discussion and support decisionmaking within the elds of transport planning and land-use planning, particularly where government contexts call for more sustainable transport options to be developed. In this respect, the
development of SNAPTA has been closely linked to the policy needs
arising from the Edinburgh Local Transport Strategy (20072012)
and subsequent reviews. Since such strategies present key sustainable transport ideas such as plans to boost transport and land-use

integration and increase the reliance on public transport, SNAPTA

provides an opportunity to deliver key elements of this strategy
so that policy decisions are based on evidence of the impacts on
accessibility. For example, using before-and-after analysis of network accessibility, SNAPTA helps to identify which centres need
to be improved or where to promote the public transport network
based on the criteria of accessibility measurement. The analysis
output prompts practitioners and decision-makers to arrange the
list of priorities and rethink the land-use patterns in locations with
high public transport accessibility. The evaluation of spatial equity
is another issue in which the application of SNAPTA can assist by
highlighting the disadvantaged parts of Edinburgh where the residents do not enjoy equal access to opportunities (i.e. areas which
require their residents to travel excessively to pursue the same
amount and quality of a particular activity when compared with
other areas around the city). Therefore, SNAPTA shows how transport and land-use integration can be clearly and visually communicated, and in so doing how the models outputs can be used to
inuence CECs transport and land-use decisions.
SNAPTA relies on a package of three accessibility measures with
a different theoretical basis and criteria to quantify the spatial
accessibility by different types of public transport modes to different types of activity opportunities, as follows:
(a) Access time to city centre. Calculating travel time or generalised cost between zones and the Central Business District
(CBD) using public transport.
(b) A contour measure. The measure describes the total number
or size of destinations that could be reached by public transport within a specic travel time. The outcomes can be
expressed either by quantity or oor space area of opportunities or economic activities which makes the measure
results simply interpreted. Different cut-off values for travel
time have been used in the analysis according to the selected
trip purpose.
(c) A potential accessibility measure. This measure is a gravitybased measure that includes a transport element, mainly
the travel time between zones, and a land-use element
determined by the quantity or size of opportunities per destination zone. A potential accessibility measure overcomes
some of the methodological limitations of a contour measure. It uses an impedance function for travel distance, time
or cost, reecting the declining attractiveness of activities at
a destination with increasing travel time (or distance) from
the origin of the journey. However, the expression of the
measure results in units that makes it less easy than the
other two measures to communicate and interpret by nonmodellers. The potential accessibility for the residents of
each origin zone (Ai) can be dened by using Hansens equation (1959), as follows:

Ai

X
aj :f t ij
J

where ai is the attractiveness (i.e. quantity or size of opportunities)

of destination zone j, tij is travel time, cost or distance from zone i to
zone j, and f (tij) is an impedance function.
Several methods have been used to estimate impedance functions in accessibility studies (see Geurs and Ritsema van Eck
(2001) for a discussion of these). This study uses a negative exponential function as the impedance function that can be expressed
in the following equation:

f t ij ebtij
where b is a sensitivity parameter to travel time. With values ranging from 0 to 1, b reduces or increases the effect of travel time

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

changes and determines the weighting of activity opportunities.

Since this study measures accessibility at a local administrative level with a high spatial disaggregation (and relatively small zones),
focusing only on use of public transport in which people are not
very sensitive to a small variation of time (Boucq, 2007; Spiekermann and Wegener, 2007), a low value of 0.1 has been selected
for b.
SNAPTA, therefore, takes into account the land use and transport characteristics of urban interactions and the availability of
opportunities which can be accessed by public transport. It focuses
on groups of people, and assumes that they have a set of social and
economic activity needs to be met at different destinations, and
that travel demand will be determined by the attractiveness of
these locations and the quality of the transport infrastructure linking these places. Issues concerning the spatial equity of public
facilities, the accessibility to workplaces and shops by public transport, and the changes to accessibility brought about by new transport infrastructure or the re-location of public facilities can all be
interrogated through the model.
The use of the measures above for SNAPTA provides a package
of accessibility measures that practitioners and decision makers
can select. These measures have been widely used in the literature
for diverse types of applications. They assess accessibility relying
on different methodologies with different levels of complexity.
Since each methodology is characterised by its own features to reect various aspects of transport and land-use systems differently,
the model users can set up the measurement framework in a way
that serves the circumstances and objectives of different applications in planning practice and satises the priority of the aspects
which must be covered. The fundamental difference between them
is that the time access to city centre and contour measures focus on
the separation between locations while the potential measure
focuses on the interaction between locations (Gutirrez et al.,
1996). The theoretical underpinnings of the potential accessibility
measure are that the interactions between an origin and destination will decline with increasing distance and time but that interactions are positively associated with the amount of activity at
each location (Hansen, 1959).
5. Methodology of accessibility modelling
The modelling approach involves the development of the following scenarios that cover the key public transport projects programmed for Edinburghs network (i.e. the tram system and reopening ESSR) within different time frames:
(1) Scenario A the base year 2011, reecting the situation of
Edinburghs transport network in 2011.
(2) Scenario B the year 2014, reecting Edinburghs transport
network after the construction of part of Phase 1a (a single
line running from the airport to the city centre).
(3) Scenario C long term development, reecting Edinburghs
transport network after the consideration of all tram lines
including those envisaged or programmed for the long term,
as follows:
Tram Phase 1a, the complete phase from and to Newhaven
and Edinburgh Airport via Haymarket.
Tram Phases 1b, from and to Haymarket and Granton
Square.
Tram Phase 2, from and to Granton Square and Newhaven.
Tram Phase 3, from and to Edinburgh Airport and
Newbridge.
Tram Line Three to South East Edinburgh, from and to Haymarket and Newcraighall.

(4) Scenario D long term development, reecting Edinburghs

transport network after taking account of all the tram lines
considered in Scenario C as well as re-opening ESSR (from
and to Waverley and Niddrie via Haymarket including eight
stations).
Six types of activity opportunities were selected to measure the
accessibility to their locations by public transport within the Edinburgh Councils area. These are: (1) the central business district
(CBD); (2) employment; (3) retail opportunities; (4) education
opportunities; (5) health opportunities; and (6) leisure and recreation opportunities. SNAPTA uses the Scottish Census Data Zones
(549 zones in Edinburgh Councils area) which are the key smallarea statistical geography in Scotland based on 2001 Census with
population between 500 and 1000 residents each (Scottish Executive, 2006b), so that contextual data on the population and socioeconomic criteria can be used. The measurement assumes that
all people living within a zone have the same level of accessibility
regardless of their different travel demands.
The location and attributes of activity opportunities have been
modelled in GIS (ARC/INFO). Land-use and socio-demographic data
(at Data Zone level) including the total number of jobs, the oor
space area of retail services and recreation facilities, and the number of patients in health care centres and hospitals, have been obtained under licence from the relevant government organisations.
The data on the number of students in secondary schools and universities, and number of leisure and recreation facilities have been
obtained from these organisations websites. Once the required
data are collected for each zone, they are linked to the associated
centroids of zones within the GIS database. Since the model assumes that all individuals are gathered in the centroids where their
journeys start and end, the determination of centroids are re-calculated on the basis of population density rather than geometric centres to avoid assigning population on non-residential areas such as
parks and large unoccupied lands. However, in this study the
accessibility impact of new transport interventions has been isolated from changes in the land-use system by xing the data on
activity opportunities in such a way that each zone holds the values of baseline year data on population, employment, retail, health,
education, and recreation in all the scenarios.
A digital multimodal transport network of bus services, tramways and ESSR railways has been built in GIS. The network covers
the whole area of study and consists of links and nodes. The nodes
are chosen on the network to correspond to bus and tram stops and
railway stations across the modelled area. For each transport link
in the GIS data base, tabular attributes of its type, length and the
time needed to pass that link have been built. SNAPTA takes into
account walk access time from the origin, waiting time, in-vehicle
time, interchange time and walk time to the destination.
Walk time is calculated as a constant multiplied by the straightline distance from the origin (i.e. the centroid of origin zone) to the
nearest public transport stop, from the disembark stop to the interchange stop, and from the nal disembark stop to the destination
(i.e. the centroid of destination zone). The calculation considers access to public transport services and interchange where the distance to a stop (or between stops) does not exceed 500 m, which
is the maximum value of the range of 300500 m walk dened
by the Scottish Transport Appraisal Guidance (STAG) (Scottish
Executive, 2003) as indicative criteria for an acceptable walking
distance to bus stops in urban areas. SNAPTA uses the value of
1.2 as a constant multiplier for the straight-line distance in Edinburgh Councils area. This value is typically applied by the City of
Edinburgh Council as a reasonable multiplier (personal communication with CEC). It is estimated based on the network patterns of

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

Fig. 4. Map showing how a multiplier of 1.2 is estimated based on a few example points around the city of Edinburgh with 800 m actual distance and 670 m radius circles.
Source: City of Edinburgh Council, Services for Communities.

several example points around the study area with the 800 m actual distance and 670 m radius circles. Fig. 4 shows the location
of six example points which have been selected randomly to estimate the multiplier value. Once walking distances are estimated,
the model uses a walk speed value of 3 mph (or 4.83 kph) for average population to measure walking time (Jones et al., 2005 and
Transport for London, 2010).
For the perceived walking time, the physical features that delay
walk access from and to public transport facilities in the beginning
and end of a journey are taken into account by estimating an extra
walk time for each zone as a weighting value of walking time. This
has been applied to slopes (e.g. for walking up a steep hill) and streets
with heavy trafc volume which causes the delay before crossing.
The total weighting value given to each zone is obtained by adding
the slope weight up to 2 min to the trafc weight up to 2 min, meaning that the maximum extra walk time for each zone is 4 min.
Wait time at the stop of origin or interchange stop is calculated
based on the minimum average of scheduled waiting time for the
selected public transport service. For example, in the case of Edinburgh bus services, wait time is calculated using the scheduled
waiting time for a service running every 10 min, since the most
regular bus services in Edinburgh run with a frequency of 6 buses
per hour during the morning peak time. This makes the minimum
average of scheduled waiting time 5 min (0.5 * 60/frequency per
hour) which is actually achieved by many services in the morning
peak time. However, the trip calculations could also be performed
with minimum wait time at the stop of origin (zero minutes),
which occurs when an individual walks to the stop at precisely
the time a bus/tram/train arrives.
The in-vehicle travel time of the currently running public transport services is calculated based on the timetables associated with
the bus and tram stops or railway stations during the morning
peak times, which already takes into account delay on the roads
because of trafc congestion. The timetables of proposed services,
particularly those for long-term development, are not all available
at the time of analysis. In this case, travel time has been estimated
based on the average time that a currently running service requires

to pass through the same route or through another route which has
the same speed limit and similar trafc volume.
Using the access time to city centre measure, accessibility is calculated based on the shortest journey time (or the fastest possible
route) during the morning peak hours by public transport from the
nearest node (bus stop, tram stop or railway station) in the network to the population-weighted centroid of each zone to the
nearest node to the centroid of the CBD. The shortest possible journey time might be achieved by using one service only or though an
interchange (one or more) between different services whether
those services are provided by the same or different operators
(i.e. Lothian Buses, FirstGroup Bus, E&M Horsburgh, Stagecoach
Bus, Edinburgh Coach Lines) with the same or different transport
mode (bus, tram or train).
The calculation of the potential accessibility measure is more
complicated. It also involves the shortest possible journey times
between any two zones using public transport. This generates a
number of relationships for each type of opportunity which is
equal to the number of origins multiplied by the number of destinations. Creating an origindestination (OD) Cost Matrix is the
technique that has been used in GIS to carry out the calculation
of the shortest journey times on the network between zones. Once
the travel time is computed for each relationship, the potential
accessibility for the residents of each origin zone is obtained by
applying Hansens equation. A contour measure has been measured for each zone by calculating the size of the desired opportunity (land use attractiveness) that can be reached by using public
transport from that node in the network nearest to the zone centroid within the specied cut-off travel time for the selected journey purpose. The study applied a cut-off value of 30 min for
travelling to a large supermarket (for food shopping) and GP practice. A length of 40 min is applied to journeys for the purposes of
work, shopping, secondary schools and leisure activities while
60 min is used for travelling to hospital and further and higher
education institutions. These values have been identied by the
Department for Transport (DfT) (2006) as the core accessibility
indicators for the key public transport journey purposes. The

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

Fig. 5. Scenario A (baseline year 2011): accessibility to jobs (based on potential accessibility measure).

Fig. 6. Relative change (improvement) in potential accessibility to jobs between Scenario A and Scenario B.

variety in the cut-off values among different journey purposes can

be explained by the fact that the choice of a supermarket and a GP
practice is not as signicant as the choice of leisure and education
facilities. Closest Facility is the GIS technique implemented to execute this measure.
6. Findings
Once the calculations have been carried out, a simulation of
the spatial distribution of accessibility is mapped in the GIS

environment based on the sum of accessibility values that

are generated for each zone acting as origin-location. Values
of the absolute and relative (percentage) changes in accessibility between the baseline scenario and the development
scenarios are computed to nd out and demonstrate the contribution of the programmed transport infrastructure to the
change in accessibility pattern to a particular activity across
Edinburgh. Also, this allows a comparison of how the different
measures incorporated in the model capture the accessibility
changes.

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

Fig. 7. Relative change (improvement) in potential accessibility to jobs between Scenario A and Scenario C.

Fig. 8. Relative change (improvement) in potential accessibility to jobs between Scenario A and Scenario D.

The results of the accessibility analyses demonstrate interesting

issues about the distribution of the impact of the tram and ESSR in
Edinburgh. Due to space constraints, the results presented in this
paper focus on the changes brought to accessibility to jobs only.
The potential accessibility measure has been used taking into account the total number of jobs based in each zone as the attractiveness of destinations. It is not surprising that the analysis results of
the current situation of the baseline scenario (see Fig. 5) and those
of the development scenarios suggest that the zones with the best
potential public transport accessibility to jobs during the morning

peak time are located in the central area of the city followed by
South Gyle Business Park in the west of Edinburgh. The map of relative change in accessibility to jobs between the baseline scenario
and Scenario B after the completion of part of tram Phase 1a scheduled for 2014 shows that the spatial variation in accessibility is
fairly small (Fig. 6). It indicates that the construction of the tram
line from the Airport to the city centre has an insignicant impact
on the potential interaction between most areas in the city for trips
to the workplace, but has greatly improved the accessibility of
some locations along the line by up to 53%.

10

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

According to Fig. 7, it is clear that the completion of all the

tram lines considered in Scenario C has an impact on the accessibility of a larger area. The accessibility to jobs of the areas where
the tram Phase 1b and the remaining part of Phase 1a are
planned to run to in a loop around the northern suburbs, connecting the city centre with the Waterfront development site, has
improved on average up to 5% with some locations having
improvements of up to 25%. Similarly, in the south east of the
city, Tram Line Three will improve the job accessibility of most
residents by up to 5%.
A comparison between Scenario C and Scenario D identies that
overall, the predicted improvement on potential accessibility to
jobs brought by the introduction of ESSR service to public transport
network is signicant. Fig. 8 shows that a considerable part of
Edinburgh, particularly the south west of the city, would benet
by running ESSR with an accessibility increase of up to 25%. Moreover, the results highlight a substantial increase in the accessibility
level of some zones around the city by up to 83% when all the tram
lines and ESSR are implemented.

7. Conclusion
The model developed in this study is not intended to provide
the complete picture of transport accessibility but it attempts to
cover adequately the required aspects of accessibility measurement and respond to some common limitations in other models
without making it very difcult to operate, interpret and, consequently, apply in practice. The challenge is not to argue that all
the gaps addressed in SNAPTA are neglected in other existing
models but it is more about building a practical accessibility
model that could offer a balance between the ease of interpretation and operationalisation and the complexity of the theoretical
basis and data disaggregation. The model has been tested and applied to the Edinburgh transport network, addressing the impact
of the tram and ESSR on accessibility to different activities at a
high level of spatial and data disaggregation of the land-use
system.
The empirical conclusion obtained in this study has demonstrated the changes in potential public transport accessibility from
the 2011 baseline case and through three different scenarios to the
completion of the full infrastructure improvements identied in
the Local Transport Strategy. GIS has been used to visualize the different types of data sets in map form portraying spacetime accessibility to services and identifying the hotspots of unequal
access. Whilst the current analysis provides information about
the changes in accessibility between the 549 data zones, it cannot
infer whether travellers perceptions of the ease of reaching the
facilities and services they require on a daily or weekly basis by
public transport will also change.
The study has not looked into the factors central to understanding modal choice, which include cultural attitudes to specic transport modes and factors associated with gender, age,
income and the number of hours spent working that inuence
travel behaviour (Weber, 2006). Although SNAPTA has been
developed with a focus on public transport modes only which is
considered as a serious limitation for some objectives, the model
has the potential to include car-based modes as well. Therefore,
further research will focus on enhancing SNAPTA by including
accessibility by private car through building the road network
taking into account the driving directions and the associated
speed limits. In addition, the model will be expanded to cover a
wider geographical area to assess accessibility and connections
between the city of Edinburgh and the key destinations in the
surrounding region such as major employment centres, universities and hospitals.

Acknowledgements
We are grateful to Damascus University, Syria which has nanced this research and to participation in the COST TU1002 project. We also want to express our gratitude to the two reviewers for
the comments and suggestions which have helped us to improve
the paper.
References
Atkins Transport Planning, 2004. Review and Options Analysis of Edinburgh South
Suburban Railway Final Report. CEC, Edinburgh.
Banister, D., Hickman, R., 2007. How to Design a more Sustainable and Fairer Built
Environment (JU5): Transport and Communications. Report to Foresight
Intelligent, Ofce of Science and Technology, London.
Boucq, E., 2007. The Effects of Accessibility Gains on Residential Property Values in
Urban Areas: The Example of the T2 Tramway in the Hauts-de-Seine
Department, France, paper presented at the European Transport Conference,
1719 October 2007. Leeuwenhorst, Netherlands.
Bowling, A., Windsor, J., 2001. Towards the good life: a population survey of
dimensions of quality of life. J. Happiness Stud. 2, 5581.
Bramley, G., Simmonds, D., Dobson, A., 2011. Using land use/transport interaction
(LUTI) models to assess the sustainability of strategic planning and
infrastructure proposals: the potential scope and illustrative applications. In:
Hull, A.D., Alexander, E.R., Khakee, A., Woltjer, J. (Eds.), Evaluation for
Participation and Sustainability in Planning. Routledge, London.
CEC, 2006. Tram Facts 8. City of Edinburgh Council, Edinburgh.
CEC, 2007. Local Transport Strategy 20072012. The City of Edinburgh Council,
Edinburgh.
CEC, 2008. Spatial Development Patterns and Urban Structure, Edinburghs
Environment: State of the Environment Audit. City of Edinburgh Council,
Edinburgh.
CEC, 2010. Transport 2030 Vision. City of Edinburgh Council, Edinburgh.
CEC, 2011. Route map, Edinburghtrams.com. City of Edinburgh Council,
Edinburgh (accessed 02.09.11).
CEC, 2013a. Public and Accessible Transport Action Plan 20132020 Consultative
Draft. City of Edinburgh Council, Edinburgh.
CEC, 2013b. Developing the New Local Transport Strategy: Issues for Review. City of
Edinburgh Council, Edinburgh.
Cervero, R., Rood, T., Appleyard, B., 1995. Job Accessibility as a Performance
Indicator: An Analysis of Trends and Their Social Policy Implications in the San
Fransisco Bay Area. Working Paper UCTC No. 366. The University of California
Transportation Center. University of California, Berkeley.
Chapin, F.S., Kaiser, E.J., 1979. Urban Land Use Planning, third ed. University of
Illinois Press, Urbana, Chicago, London.
MVA Consultancy, 2008. Accessibility Analysis in the SEStran Area 2007/08.
SEStran, Edinburgh.
Curtis, C., Scheurer, J., 2010. Planning for sustainable accessibility: developing tools
to aid discussion and decision-making. Prog. Plan. 74 (2), 53106.
Department for Transport, 2006. Guidance on Accessibility Planning in Local
Transport Plans. DfT, London.
Forward, S., 2003. State of the Art Report on Life Quality Assessment in the Field of
Transport and Mobility. Swedish National Road and Transport Research
Institute, Linkoeping, Sweden.
Geurs, K.T., Ritsema van Eck, J.R., 2001. Accessibility Measures: Review and
Applications. National Institute of Public Health and the Environment,
Bilthoven.
Geurs, K.T., van Wee, B., 2004. Accessibility evaluation of land-use and transport
strategies: review and research directions. J. Transport Geogr. 12, 127140.
Gutirrez, J., Gmez, G., 1999. The impact of orbital motorways on intra-metropolitan
accessibility: the case of Madrids M-40. J. Transport Geogr. 7 (1), 115.
Gutirrez, J., Gonzlez, R., Gmez, G., 1996. The European high-speed train network:
predicted effects on accessibility patterns. J. Transport Geogr. 4 (4), 227238.
Halden, D., 2002. Using accessibility measures to integrate land use and transport
policy in Edinburgh and the Lothians. Transport Policy 9, 313324.
Halden, D., McGuigan, D., Nisbet, A., McKinnon, A., 2000. Accessibility: Review of
Measuring Techniques and their Application. Scottish Executive Central
Research Unit, Edinburgh.
Handy, S.L., Niemeier, D.A., 1997. Measuring accessibility: an exploration of issues
and alternatives. Environ. Plan. A 29, 11751194.
Hansen, W.G., 1959. How accessibility shapes land use. J. Am. Inst. Plan. 25, 7376.
Himanen, V., Lee-Gosselin, M., Perrels, A., 2005. Sustainability and the interactions
between external effects of transport. J. Transport Geogr. 13, 2328.
Holl, A., 2006. A review of the rm-level role of transport infrastructure with
implications for transport project. J. Plan. Literat. 21, 316.
Hull, A.D., Karou, S., 2011. Analysis of the Impact of Change in Transport Delivery on
Accessibility: The Case of Edinburgh Public Transport, paper presented to the
World Planning Schools Congress, Perth, July 2011.
Jones, P., Wixey, S., Titheridge, H., Christodoulou, G., 2005. Developing Accessibility
Planning Tools. Transport Studies Group University of Westminster.
Karou, S., Hull, A.D., 2012. Accessibility measures and instruments. In: Hull, A.D.,
Silva, C., Bertolini, L. (Eds.), COST Action TU1002 Accessibility Instruments for
Planning Practice. COST Ofce, Porto, pp. 119.

S. Karou, A. Hull / Journal of Transport Geography 35 (2014) 111

Keller, J., Arce-Ruiz, R., Condeo-Melhorado, A., Mavridou, M., Nordstrm, T., Ortega,
E., Tenny, A., Trova, V., Hull, A.D., 2012. Accessibility in planning practice. In:
Hull, A.D., Silva, C., Bertolini, L. (Eds.), COST Action TU1002 Accessibility
Instruments for Planning Practice. COST Ofce, Porto, pp. 2143.
Koenig, J.G., 1980. Indicators of urban accessibility: theory and application.
Transportation 9, 145172.
Leitmann, J., 1999. Can city QoL indicators be objective and relevant? Towards a tool
for sustaining urban development. In: Lim, L.Y., Yuen, B., Low, C. (Eds.), Urban
Quality of Life: Critical Issues and Options. School of Building and Real Estate,
National University of Singapore.
Morris, J.M., Dumble, P.L., Wigan, M.R., 1979. Accessibility Indicators for Transport
Planning. Report No. 102. Australian Road Research Board, Vermont South.
NICHES, 2007. Facilitating Urban Transport Innovation on the European Level.
Research and Policy Recommendations, <http://www.niches-transport.org>
(accessed).
OSullivan, P., 1980. Transport Policy. An Interdisciplinary Approach. Batsford
Academic and Educational Ltd., London.
Priemus, H., Nijkamp, P., Banister, D., 2001. Mobility and spatial dynamics: an
uneasy relationship. J. Transport Geogr. 9, 167171.
Reneland, M., 1998. Begreppet tillgnglighet, GIS-projektet Tillgnglighet i
svenska stder 1980 och 1995. Chalmers University of Technology,
Gothenburg. (The concept of accessibility. The GIS project: Accessibility in
Swedish Towns 1980 and 1995).
Roseland, M., 1997. Dimensions of the eco-city. Cities 14 (4), 197202.
Ross, W., 2000. Mobility and accessibility; the yin and yang of planning. World
Transport Policy Pract. 6 (2), 1319.
Scottish Executive, 2003. Scottish Transport Appraisal Guidance (STAG). Scottish
Executive, Edinburgh.

11

Scottish Executive, 2004. Scotlands Transport Future: The Transport White Paper.
Scottish Executive, Edinburgh.
Scottish Executive, 2006a. Scotlands Transport Strategy. Scottish Executive,
Edinburgh.
Scottish Executive, 2006b. Scottish Neighbourhood Statistics Guide. Scottish
Executive, Edinburgh.
Spiekermann, K., Wegener, M., 2007. Update of Selected Potential Accessibility
Indicators. Final Report, Urban and Regional Research (S&W) and RRG Spatial
Planning and Geoinformation.
Sultana, S., 2006. What about dual-earner households in jobs-housing balance
research? An essential issue in transport geography. J. Transport Geogr. 14,
393395.
Transform Scotland, 2007. Reopening the South Sub brieng, <http://
www.transformscotland.org.uk> (accessed).
Transport for London, 2010. Measuring Public Transport Accessibility Levels
PTALs: Summary. Transport for London, London.
Transport Initiatives Edinburgh, 2004. Integrated Transport Initiative Proposed
Congestion Charging Scheme Statement of Reasons Final Draft Order, TIE,
Edinburgh.
Transport Initiatives Edinburgh, 2006. Edinburgh Tram Network: Draft Final
Business Case. TIE, Edinburgh.
Transport Initiatives Edinburgh, 2007. Edinburgh Tram Network: Final Business
Case. TIE, Edinburgh.
Weber, J., 2006. Reections on the future of accessibility. J. Transport Geogr. 14,
399400.
Yigitcanlar, Tan, Sipe, Neil G., Evans, Rick, Pitot, Matt, 2007. A GIS-based land use
and public transport accessibility indexing model. Austr. Plan. 44 (3), 3037.