ADDIS ABABA UNIVERSITY

ADDIS ABABA INSTITUTE OF TECHNOLOGY

SCHOOL OF CIVIL AND ENVIROMENTAL ENGINEERING

Accessibility and Suitability Analysis of Light Rail Station

Location by using (AHP) and GIS: Case study on Existing
and Future Expansion of Addis Ababa LRT Respectively

A thesis in Railway Engineering Center

By: Gomeju Taye

Dec, 2016

A thesis submitted to the school of Graduate Studies of Addis Ababa

University in Partial fulfillment of the degree of Master of Science in Civil
Engineering
Acknowledgements
Foremost, I would like to thank my Almighty GOD for saving me to this piece of second and
for his help in all my way here and next, I wish to express my gratitude to all those who have
given me assistance, help and support during the completion of my Masters of Science in
Division of Railway Engineering, Graduate School of Civil and Environmental Engineering
at Addis Ababa University.
Persistently, I would like to express my deep and sincere gratitude to my academic advisor,
Dr. Alemayehu Ambo, for his continuous support for successfully completing my thesis, and
for his kind assistance, insightful comments and knowledge that added considerably to my
academic experience. I appreciate his assistance in this matter. My research would not have
succeeded without his helps and support.
Finally, I would like to consider here the continuous encouragement and moral support I
received from every member of my family throughout my studies and appreciation to my
parent who have supported me in everything I have undertaken and whose love has
accompanied me throughout my studies. I would also like to thank my older brothers and my
younger sister for their concrete help, my close friends in AACP, motivated and encouraged
me when I felt uninspired and for this, they have earned my deepest gratitude.

i
Abstract
Addis Ababa city metropolitan constructed phase one light rail transit line aiming to solve
current transport problem. This Light Rail Transit is assumed to transport 80,000 passengers
per hour per direction. For the system to hit the targeted objective on the proposed line,
stations have to be positioned at a place where it can attract maximum users which enable the
system to solve the problem and to be independent of subsidies. This is all about integrating
Light rail system with other transport modes in Addis Ababa.

This paper examines the existing station sites accessibility to other transportation modes or
access mode, and adopts Geographic Information Systems and Multi-Criteria Evaluation
technique to carry out suitability mapping of station locations for phase two light rail transit
route in Addis Ababa. For the station suitability analysis, eleven different criteria were
identified and each criterion was weighted using Analytical Hierarchy Process AHP. The
output of Analytical hierarchy process was used as an input for geographic information
system GIS special analysis. Finally, based on these criteria requirements using overlaying,
Euclidean distance calculation, Rasterisation, buffering reclassification and weighted overlay
analysis, the station site suitability map was generated. The map revealed five classifications
as: “less suitable”, “suitable”, “moderately suitable”, highly suitable and “extremely suitable”.

Key words: Accessibility, AHP, GIS, LRT, Station, Suitability and Super Decision.

ii
 Table of Contents
Acknowledgements ..................................................................................................................... i
Abstract ......................................................................................................................................ii
Table of Contents ......................................................................................................................iii
Table of Tables ......................................................................................................................... vi
Table of Figures .......................................................................................................................vii
List of Acronyms ....................................................................................................................viii
CHAPTER 1 .............................................................................................................................. 1
1 INTRODUCTION .......................................................................................................... 1
1.1 General .................................................................................................................... 1
1.2 Why Analytic Hierarchy Process (AHP) ................................................................. 5
1.3 Research Identification ............................................................................................ 5
1.3.1 Research Problem............................................................................................. 5
1.3.2 Research Objectives ......................................................................................... 6
1.3.3 Research Questions .......................................................................................... 6
1.3.4 Conceptual Framework .................................................................................... 7
1.3.5 Thesis Organization ......................................................................................... 7
1.3.6 Research Design ............................................................................................... 8
CHAPTER 2 .............................................................................................................................. 9
2 LITERATURE REVIEW ............................................................................................... 9
2.1 General .................................................................................................................... 9
2.2 Introduction ............................................................................................................. 9
2.3 Light Rail Transit System ........................................................................................ 9
2.4 Light Rail Transit (LRT) Station ........................................................................... 11
2.4.1 Definition ....................................................................................................... 11
2.4.2 Types of stations ............................................................................................ 13
2.4.3 Station Location Planning/selection ............................................................... 15
2.4.4 Station location selection criteria ................................................................... 16
2.4.5 Station accessibility evaluation ...................................................................... 18
2.5 LRT Station location planning using AHP and GIS ............................................. 23
2.5.1 Analytical Hierarchy Process (AHP) ............................................................. 23
2.5.2 Geographic information system (GIS) ........................................................... 26
2.5.3 Application of GIS in Railway Engineering .................................................. 26
CHAPTER 3 ............................................................................................................................ 28
3 METHODOLOGY ....................................................................................................... 28

iii
 3.1 Introduction ........................................................................................................... 28
3.2 Methods Used in New Station Accessibility Analysis .......................................... 29
3.3 Methods Used in Station Suitable site Selection ................................................... 29
3.3.1 Euclidean Distance Calculation ..................................................................... 30
3.3.2 Buffering ........................................................................................................ 30
3.3.3 Rasterisation ................................................................................................... 31
3.3.4 Reclassification .............................................................................................. 31
CHAPTER 4 ............................................................................................................................ 32
4 DATA COLLECTION ................................................................................................. 32
4.1 Study Area ............................................................................................................. 32
4.2 Primary Data .......................................................................................................... 35
4.2.1 General ........................................................................................................... 35
4.2.2 Participants ..................................................................................................... 35
4.3 Secondary Data ...................................................................................................... 35
CHAPTER 5 ............................................................................................................................ 37
5 PHASE ONE STATION ACCESSIBILITY ANALYSIS ........................................... 37
5.1 Station Access Guidelines and Standards.............................................................. 37
5.1.1 Pedestrian Access Guidelines and Standards ................................................. 38
5.1.2 Bus Transfer Access Guidelines and Standards ............................................. 38
5.1.3 Bicycle Access Guidelines and Standards ..................................................... 39
5.1.4 Auto Access ................................................................................................... 39
5.1.5 Kiss-and-Ride ................................................................................................. 39
5.2 Results and Discussions ........................................................................................ 41
5.2.1 Criteria Mapping and Analysis ...................................................................... 41
CHAPTER 6 ............................................................................................................................ 54
6 PHASE II STATION LOCATION SUITABILITY ANALYSIS ................................ 54
6.1 Development of Station Location Selection Criteria ............................................. 54
6.1.1 Classification of criteria maps and data analysis ........................................... 55
6.1.2 AHP Model Development .............................................................................. 61
6.1.3 Criteria Weights and Class Weights .............................................................. 62
6.1.4 Pair-wise Comparisons................................................................................... 62
6.1.5 Factors pair-wise comparison ........................................................................ 63
6.1.6 Weighting Criteria .......................................................................................... 65
6.1.7 Consistency Check ......................................................................................... 66
6.2 Results and Discussions ........................................................................................ 67
6.2.1 Multi-criteria preparation for GIS .................................................................. 67
6.2.2 Geodatabase ................................................................................................... 67

iv
 6.2.3 Euclidean Distance Calculation ..................................................................... 68
6.2.4 Buffering ........................................................................................................ 68
6.2.5 Rasterisation ................................................................................................... 68
6.2.6 Reclassification .............................................................................................. 69
6.3 Criteria Mapping and Spatial Analysis .................................................................. 69
6.3.1 Distance from Road........................................................................................ 69
6.3.2 Slope............................................................................................................... 70
6.3.3 Land-use ......................................................................................................... 71
6.3.4 Population ...................................................................................................... 72
6.3.5 Proximity to bus stops .................................................................................... 74
6.3.6 Proximity to parking building ........................................................................ 75
6.3.7 Weighted Overlay .......................................................................................... 75
6.4 Suitability Map ...................................................................................................... 76
6.5 Suitable Location Identification ............................................................................ 77
CHAPTER 7 ............................................................................................................................ 84
7 CONCLUSIONS AND RECOMMENDATIONS ....................................................... 84
7.1 Conclusions ........................................................................................................... 84
7.2 Recommendations ................................................................................................. 85
CHAPTER 8 ............................................................................................................................ 86
8 FUTURE RESEARCH AREAS ................................................................................... 86
Reference ................................................................................................................................. 87
Appendix .................................................................................................................................. 90
Appendix 1 ........................................................................................................................... 90
Appendix 2 ........................................................................................................................... 94
Appendix 3 ........................................................................................................................... 94
Appendix 4:.......................................................................................................................... 95
Appendix 5 ........................................................................................................................... 96

v
 Table of Tables
Table 4.1: Required Data ......................................................................................................... 36
Table 5.1: shows Rail-Bus Transfer Distance Standards ........................................................ 39
Table 5.2: Summary of study area stations accessibility ......................................................... 51
Table 6.1: Criteria Developed for Station Location................................................................. 55
Table 6.2: Relative Importance of each criterion.................................................................... 63
Table 6.3: Saaty nine point scale ............................................................................................. 63
Table 6.4: Five-point scale for pair wise comparison of sub-criteria scoring: ........................ 63
Table 6.5: Category, land use, score and class ....................................................................... 64
Table 6.6: Category, slope, score and class ............................................................................ 64
Table 6.7: Pair-wise Comparison of Distance from road factor .............................................. 64
Table 6.8: Pair-wise Comparison of Distance from LRT Line factor ..................................... 64
Table 6.9: Pair-wise comparison of proximity to Parking Building Factor ............................. 64
Table 6.10: Pair-wise comparison of proximity to Bus Station Factor.................................... 65
Table 6.11: Calculate criteria weights .................................................................................... 65
Table 6.12: Consistency Index value ....................................................................................... 67

vi
 Table of Figures
Figure 1.1: Conceptual frameworks ........................................................................................... 7
Figure 1.2: Research Structure ................................................................................................... 8
Figure 2.1: Traffic and land-use interaction (traffic spiral) ..................................................... 11
Figure 2.2: Stations Based on Their Location Relative to Tracks (Online, Inline and Offline)
.......................................................................................................................................... 14
Figure 2.3: Station design checklist ......................................................................................... 19
Figure 3.1: Methods Flow Chart Used in Station Access Evaluation...................................... 29
Figure 3.2: Methods Flow Chart in Suitable Station Site Analysis ......................................... 30
Figure 4.1: Addis Ababa existing and future extension LRT network .................................... 33
Figure 4.2: Study Area for existing stations accessibility analysis .......................................... 34
Figure 4.3: Addis Ababa Street Network Converted to GIS shape file and Population Density
.......................................................................................................................................... 36
Figure 5.1: Typical Intermodal Railway Station Interface with Bus ....................................... 40
Figure 5.2: Study area LRT station connectivity with other modes ........................................ 42
Figure5.3: Proposed City Center Parking Building Multiple-Ring Buffering ......................... 44
Figure 5.4: Park and Ride Locations in Addis Ababa ............................................................. 45
Figure 5.5: 2010 Addis Ababa modal shares .......................................................................... 47
Figure 5.6: Circular station service area within catchments of 500m radius ........................... 48
Figure 5.7: Connectivity to Bus top, park-n-ride and taxi stops .............................................. 50
Figure 5.8: Study area Population density by kebeles ............................................................. 52
Figure 5.9: Study area Land use and LRT stations .................................................................. 53
Figure 6.1: Addis Ababa Land Cover ...................................................................................... 56
Figure 6.2: Road Euclidean distance ....................................................................................... 57
Figure 6.3: Study area population density ............................................................................... 58
Figure 6.4: Slope Map of Addis Ababa ................................................................................... 59
Figure 6.5: Proximity to Parking Building .............................................................................. 60
Figure 6.6: Proximity of LRT stations to Stations ................................................................... 61
Figure 6.7: Four level AHP Model .......................................................................................... 62
Figure 6.8: Structure of the created Geodatabase .................................................................... 68
Figure 6.9: Reclassified Road Euclidean Distance .................................................................. 70
Figure 6.10: Reclassified slope ................................................................................................ 71
Figure 6.11: Reclassified Land use .......................................................................................... 72
Figure 6.12: population density reclassification window ........................................................ 73
Figure 6.13: Reclassified population Density .......................................................................... 73
Figure 6.14: Reclassified Bus Station Proximity ..................................................................... 74
Figure 6.15: Reclassified Parking Building Proximity ............................................................ 75
Figure 6.16: Weighted Overlay Window ................................................................................. 76
Figure 6.17: Model Geoprocess ............................................................................................... 77
Figure 6.18: LRT station site suitability model ....................................................................... 77
Figure 6.19: Station site suitability map for Torhaloch to Sebbeta National railway station .. 78
Figure 6.20 : Station site suitability map for route from Arada to Addisu gebeya .................. 79
Figure 6.21: Station site suitability map for route from Summit to Laga Tafo ....................... 80
Figure 6.22: Station location Suitability map for route from Kaliti to Akaki ......................... 81
Figure 6.23: general station suitability map ............................................................................. 82
Figure 6.24: Potential Intermodal Station location .................................................................. 83

vii
 List of Acronyms

AACRA Addis Ababa City Road Authority

AACPPO Addis Ababa City Planning Project Office

AHP Analytical Hierarchy Process
BRT Bus Rapid Transit
CBD Central Business District
GIS Geographic Information System
LRT Light Rail Transit
MRT Metro Transit
Park-n-ride Park and ride
TOD Transit Oriented Development
VMT Vehicle Meter Traveled
MCDM Multi-criteria decision making

viii
 CHAPTER 1
1 INTRODUCTION
1.1 General
Transport planning is crucial in planning sustainable developments and ensuring accessibility
to passengers, goods and services. One of the key elements in sustainable transport planning
is to minimize the distance passengers travel and goods hauled; and if longer distances
required, then, good network connections should be provided. Cities shall create walk-able
environment by providing pedestrian barrier free network, a cycle-friendly community and
private, public and non-motorized transport must functionally complement each other by
forming balanced intermodal systems or intermodal network.

Hence, all urban transportation components; public (mass transit) and private transport should
be integrated to easily and efficiently connect downtown urban areas with low density areas
or suburbs areas. Therefore, parking should be provided at light rail transit stations in such a
way that, buses, taxis and passenger cars should feed passengers from suburbs to downtown
light rail transit system.

As stated by Ahmad Muhammad et al, (2006), in order to bring this connection, it is

extremely important to devise an effective policy on the planning, management and operation
of the public transportation system including such as light rail transit (LRT). However, the
main problems usually identified are the aspect of integration that is, mistakes in integrating
LRT route or alignment and station or the failure in taking into account surrounding land
uses, or the linkages between other public transportation modes that are not properly planned.
The lack of integration often creates problems, such as minimum usage of the LRT services
contributing to the financial loss incurred by the rail operators which intern makes transit
subsidy dependent.

One objective of modern transportation is changing and promoting single mode usage to
intermodal transportation by combining strengths and offset weaknesses of various
transportation options. The major goal of intermodal passenger transport is to reduce
dependence on automobile as major mode of surface transportation and increase use of public
transport such as LRT and Bus Rapid Transit (BRT). To assist and encourage such trips,
Light railway transit stations have to be sited at optimum locations where it can be accessible

1
and easily integrated with other modes for passenger transfer and enhance easy walk in
traffic. To develop active urban transportation, metro, light rail transit (tram and LRT) and
bus transit are being used worldwide.

“Public transport systems are more likely to be regarded as attractive alternatives to private
car if they operate in a joined-up, integrated way. Integration involves co-ordination between
the services, physical proximity allowing ease of interchange at stations, and through
ticketing and widespread availability of passenger information about routes, fares and
timetables. Passengers consider the level of integration to be the least satisfactory aspect of
light rail system. Integration with bus services has been poor to moderate on many lines and
bus and light rail services have been in competition with one another on the same routes”
(UK Tram, 2012).

For railway industry, in order to be economically efficient and capable of handling current
and projected travel demand of the cities, routes and station locations have to be planned and
constructed based on existing and planned urban land use, inter connectivity with road
networks, opportunity it can bring to the surrounding areas, minimum right-of- way cost,
constructability and its operability.
Historically, most transportation corridors have been developed for a single mode, with the
purchase of only the right-of-way required to accommodate a particular transportation
improvement. The future emphasis in corridor planning will be multiuse and multimodal, as
new transportation modes and communication facilities are implemented in and retrofitted
into existing corridors. To optimize the limited rights-of-way available, LRT will
increasingly be designed in joint or shared use with other facilities. The emphasis will be on
person capacity per travel lane or track.

The rail transit stations are located to provide convenient linkages between users' origins and
destinations in such a way that, access and transfer time is minimized. The planning of rail
transit station location requires understanding of demographics, land use, topography of an
area. Light rail transit station should be optimally located by considering these factors, to
increase the usage of rail transit services.

A light rail transit lightweight metropolitan electric railway system is characterized by its
ability to operate single cars or short trains along exclusive right-of-way at street level. These

2
vehicles are usually powered by overhead electrical wires and offer frequent, fast, reliable,
comfortable and high quality services that is environmentally sustainable (Boorse, 2000:443).

Light rail transit is often identified by its right-of-way and vehicle weight and size; when
compared with a regional railway or metro system. However, when compared with modern
low-floor trains, LRT is heavier because the vehicles are usually wider or there are two to
three vehicles coupled together. The terms „heavy‟ or „light‟ do not solely refer to weight, but
also to the flexibility of a system to deal with different types of right-of-ways and to the
ability to be integrated into a variety or urban streetscapes (Topp, 1999).

In many developed cities today, transportation is characterized by the dominance of the car
and high auto-dependency in travel. And this causes congestion and steals billions of hours a
year.

Therefore, for a city to provide active transportation system there should be public transport
domination and good transport links. Light Rail Transit (LRT), Metro, tram and Bus Rapid
Transit (BRT) are modern public mass transport designed to resolve congestions in global
cities.
Light rail transit station location planning where it provides maximum links between
transport modes and walk-in traffic is component of active transport system planning. And
urban transport planning is a foundation for creating easily accessible and economically well-
built city. And good station planning is essential to maximize new rail line ridership and to
help provide the right amount of parking at stations along the line (Burger and Byrne, 2008).

Many researchers have conducted study on how to plan railway station location using
different methodologies. Some of which are explained as follow. Scott A. Burger and Bill
Byrne (2008) acknowledged three station location planning elements:
 Determine station Locations: Select the number of stations using technology dependent
spacing guidelines and qualitative community factors. Land availability, highway access,
proximity to key land uses, shared parking possibilities and transit oriented development
(TOD) potential are key considerations.
 Parking Supply: Obtain parking demand input from a travel demand model. Establish
the required overall corridor parking supply, considering factors such as the capital
construction budget, expected passenger trip types and station access modes. Determine
the parking distribution across the stations. Provide larger lots towards the end of the line.

3
  Assess Traffic Impacts: Determine trip production rates for parking spaces, preferably
from local experience. Define traffic impact guidelines, in conjunction with affected
agencies. Select intersections in close proximity to the station areas to assess traffic
operations and identify needed improvements.

According to Metropolitan Council (2012), station location planning elements are:

(i.) Major travel patterns (including location of major activity centers)
(ii.) Population and employment density
(iii.) Auto ownership
(iv.) Trip purpose (e.g., commuters, students, shoppers, other)
(v.) Existing transit ridership

(vi.) Commuter market analysis (geographic market area, existing and future demand, and
facility and service competition or reinforcement)

Urban railway transit line selection and station location is a challenging problem. This is
because; there are many factors that should be considered at a time to select optimum route
and station location i.e. railway alignment and station selection is multi-criteria problem.

In view of the foregoing, this research focuses on light rail transit (LRT) optimal station
location planning based on factors that can simultaneously determine the station location
because, planning of the railway station locations in integrated passenger transport is a multi-
criteria problem. Therefore, multi-criteria decision making process is integrated with GIS to
develop station site suitability map.

Multi-criteria optimization, precisely multi-criteria decision-making (MCDM) is the process

of simultaneously optimizing two or more conflicting objectives subject to certain
constraints. This can be found in various fields, for example in an engineering design as well
as in financial topics of an organization, precisely wherever optimal decisions need to be
taken in the presence of trade-offs between the conflicting objectives. In each case, we are
looking for a solution that has optimized each objective in such a way that if we try to
optimize the solution any further, then, as the result, the other objective(s) will suffer. When
setting up and solving a multi-criteria optimization problem, the goal is to find a solution, and
to quantify how much it is better than any other solution.

4
1.2 Why Analytic Hierarchy Process (AHP)
The Analytic hierarchy process (AHP) is a method that helps the decision maker in
evaluating the complicated problems and issues. Furthermore, this method contributes in
identifying numerical values for the objective stimuli related to the given problem through
conducting comparisons between the various criteria that affect the problem directly. This
method is crucial since it assists the decision makers in observing the continuous interaction
between the elements of the complicated problem. However, this in turn assists them in
defining the problem's elements and setting its priorities depending on their relevant
knowledge and experiences as well as the desired goals (Satty, 1990).

1.3 Research Identification

1.3.1 Research Problem
Addis Ababa‟s concentric land use pattern, service and administrative institutions are
concentrated at the center. This in turn resulted in traffic congestion.

Another cause of traffic congestion in Addis Ababa is the rapid urban development and high
population inflows from different part of the country for high employment opportunities as a
result of booming construction. Poor traffic management also contributes to congestion in
Addis Ababa.

Current problems of Addis Ababa City Transport are identified as follows:

• Increasing traffic congestion;
• Declining attractiveness of road-based public transport in quality and quantity;
• Increasing high costs of travel and journey times;
• High levels of road accidents;
• Due to increased road vehicles high exhaust emissions;
• The existing public transport system in Addis Ababa is critically inadequate to
provide service for the existing travel demand;
• The number of mini-bus taxis is not sufficient;
• Increasing demand for parking;
• Low affordability;
• Non-motorized transport such as carts and cycles are operating at the periphery of the
City unlicensed and without respecting traffic rules; etc. (Tefera, 2013)

5
As a result of devastated current transportation activities in the city and to lessen the problem,
light rail transit system for mass transportation is being constructed. Although LRT is a
system by itself, it is a combination of sub-systems such as, track, station and rolling stock.
Railway can be an effective tool for facilitating urban growth, if planned and designed
appropriately and properly tied in with urban development.
In order to meet the targeted aim, there should be inter-connection between LRT system and
other modes; so that, the rail transit should not result in traffic congestion. For the rail transit
not to cause traffic congestion, its stations and terminals should be located where they can be
integrated with other modes of transportation.

So in this research an attempt was made to identify optimal LRT station locations where it
can be integrated with other transport modes and surrounding land-use. This integration has a
vital role in attracting large amount of passengers and revenue for transit system.

1.3.2 Research Objectives

The general objective of this research is to develop a method that enables analysis of LRT
station site suitability and accessibility to other transportation modes users by integrating
AHP and GIS Spatial Analysis Tool.

In this regard, three sub-objectives have been arranged step by step that function
subsequently. The following are the three sub-objectives:
(i) To develop a method for LRT station location accessibility analysis.
(ii) To develop LRT station site suitability map building model

1.3.3 Research Questions

For each sub-objective, a question has been put forward to guide the research. These
questions will be answered accordingly in this thesis. Consequently, a conceptual
framework is introduced to illustrate how those questions have been organized to achieve
specific research objectives. The following are the research questions:

(i.) What factors should be considered when analyzing station site accessibility? This
helps to identify input criteria in developing a method for existing Addis Ababa LRT
station accessibility analysis.

(ii.) What criteria should be considered and how these criteria used in developing station
site suitability map? This helps to identify station location determining criteria

6
(iii.) How GIS and AHP be integrated to identify potential LRT station sites? This helps to
identify suitability map for LRT station locations.

1.3.4 Conceptual Framework

Urban light rail transit stations are gateways for travelling public to access rail services.
Passenger stations comprise functional elements that support arrivals and departures of
passengers that include: platforms, station buildings, pickup and drop off areas, and
intermodal connections. These elements have a significant impact on attracting passengers
(Northern New England Intercity Rail Initiatives, 2014). Figure 1.1 illustrates conceptual
framework.

Figure 1.1: Conceptual frameworks

1.3.5 Thesis Organization

The research is comprised of eight chapters and several sub-sections. Chapter one deals with
introduction followed by Chapter two which covers literature review; Chapter three deals
with methodology; Chapter four deals with study area background; Chapter five deals with
station accessibility analysis; Chapter six deals with station site Suitability analysis; Chapter

7
seven forwards conclusions and recommendations and finally Chapter eight presents future
research areas.

1.3.6 Research Design

The general steps of conducting this research are shown in figure1. 2.

Figure 1.2: Research Structure

8
 CHAPTER 2

2 LITERATURE REVIEW
2.1 General
The literature review chapter of this thesis consists of two sections based on the main
objective: (i) accessibility analysis of stations and; (ii) research associated with suitability
Map development for light rail transit stations. Section one inspects studies on accessibility of
rail transit station while inspects analytical methods for determining urban rail transit station
suitable sites and selection criteria.

2.2 Introduction
Locating and selection of urban rail transit station is a task that considers many factors and
different professionals such as railway engineer, urban planner, economists and stakeholders
mainly; users and governments that should come together to identify the factors and to
determine their relative importance. Therefore, locating a station is not an easy task, if we
consider locating station at optimum position to attract maximum ridership and integrate with
other transportation modes.
Accessibility of rail transit is determined by station ability to draw commuters from
catchment area of 400m-500m radius in high density down town and greater area in suburban
areas.
According to the U.S. Department of Transportation Federal Railroad Administration Office
of Railroad Policy and Development 2011: 4), the development of station area plan begins
with developing a vision, setting goals and objectives; followed by the three primary work
areas: station location, transportation connectivity, urban design and infill development.

2.3 Light Rail Transit System

Light rail transit system is a sum of different sub-systems, such as: tracks, rolling stocks
(vehicles), stations, and electrification and signals that are accurately integrated. Perrott &
Menzies (2009) broadly defined urban LRMT Light Rail Metro Transit systems as urban rail
transportation solutions that use electrically powered coaches (known as rolling stock) to
transport customers between fixed stations. According to them, Light Rail Metro Transit) has
different classifications based on system right-of-way whether they are segregated or not and
rolling stocks capacity.

9
Light rail transit is also defined by Robert R. Clark (1984), as mode of passenger
transportation using a fixed guide way system and generally employs electricity from external
source as a means of propulsion. He described LRT as a versatile mode which has difference
in performance and capacity from system to system and from line to line within the same
system that leads LRT as a viable solution to very large transportation problem.

Antero Alku (2011) stated Light rail system as a four form system and these are:
• It works in the tunnels as a metro;
• On own right of ways; it works like a train;
• On streets, it works like a traditional tramway or a bus; and
• On market squares and other pedestrian areas, it works like an old fashion slow
streetcar or coaches.

Economically, light rail transit system is cost-intensive investment and most of the time;
operation is dependent on government subsidies. In order to create subsidy-free light rail
transit system, proper planning and operation are vital instruments. Lai (2012) stated that rail
transit system has to be competitive with other transport modes and attract sufficient trips to
generate revenue and reduce congestion, in order to justify the high costs associated with
investment and operation. If it is planned properly, light rail transit investment is expected to
encourage and concentrate economic and community growth around the rail stations and
along rapid railway routes. Transit system station needs to be integrated with surrounding
land use to encourage TOD (transit oriented development).

Vuchic (2005) described the advantages of LRT over buses as spacious articulated cars
operating in 1-4 trains that may use short tunnels in city centers or operates in pedestrian
streets which makes LRT largely free from traffic congestion, allow them higher operational
speed, reliable service and strong and attractive image in the city. LRT is a transit mode that
has performance and investment cost between buses and metro systems and it has diversity in
physical and performance features that makes it applicable at locations.

The reason for LRT is that construction of large road network would not bring sustainable
active urban transportation. Since increasing road space may reduce the quality of urban
environment, prevent people from walking and cycling, and force those households who can
afford to move into cleaner and less noisy exurban areas. The interaction between transport
and land-use, and the dynamics of related developments must be considered.

10
Deutsche Gesellschaft for Tecnische Zusammenarbeit (GTZ) Gmbh (2004) explained that a
car-oriented life style is out of reach of most people in developing countries and that the
increment in private car and other motorized transport service leads to overload of existing
roads, congestion and environmental degradation of urban space.
Now days, a solution for this problem and sustainable urban transport is mass transit (LRT
and BRT) or transit oriented development (TOD) around or along transportation systems.
Only public transport can assure mobility in large cities. Figure 3 illustrates traffic and land-
use interaction (traffic spiral) as provided by Deutsche Gesellschaft for Tecnische
Zusammenarbeit (GTZ) Gmbh, (2004).

Figure 2.1: Traffic and land-use interaction (traffic spiral)

2.4 Light Rail Transit (LRT) Station

2.4.1 Definition
A station is a point along railway line where trains stop for passengers to alight and/or board
and for freight to load and/or unload. Unlike bus stops that can be easily relocated, rail transit
stations are permanent structures that involve major investment and often have strong impacts
on their surroundings (Vuchic, 2005). Light rail transit station is only for passenger traffic.
As explained in previous section, LRT station is one of the components of light rail transit
system. Many scholars and books defined LRT station in the following ways:-
11
• A station is defined as a public transport facility of high quality, which acts as a central
departure and/or destination point to accommodate high passenger volumes where
passengers transferred from and to different modes of transport services. Stations are
normally located at points of high travel demand and can be sited from outer sub-urban
areas to inner-city areas and generally serve key catchment areas such as commercial and
business districts, and may contain a moderate to a very high level of supporting
structures such as park-and-ride, kiss-and-ride, public and bicycle amenities (Queensland
Government, 2012). Stations are also considered as shopping malls; meeting places and
urban landmarks. Development of railway station attracts economic activities such as
shops, retail and offices at its surrounding. Amtrak (2013:7) defined station as a major
hub in a multimodal network, connecting downtown and other important places in the
region. Further definitions were provided as follows:

• Transfer stations are joint stations for two or more lines at which passengers between
lines, terminals are, strictly defined, end stations on a transit line, but sometimes the
term is also used for major transfer stations (http://www.amazon.com/Urban-Transit-
Operations-Planning ).

• A light rail station is a station or stop in a rapid transit system. It can be as simple as a
bus stop or as extravagant as an underground or elevated multi-use transit hub
(Topalovic et al., 2009).

• Light rail station is point where passengers board and alight from trains, and ranges
from simple platforms at ground level to complex structures above or below ground
which may be accessed via stairways, escalators and elevators (This Is Light Rail
Transit, 2000:12).

Station locations and park-and-ride considerations should be reflective of economic

development objectives in that area because stations have many advantages in developing the
surrounding areas. The following are some of the advantages:

• Stations are potential catalysts for new transit-supportive development;

• Limiting free surface parking at or near stations will make more land available near
the station for higher-density development where zoning allows;

• Station can serve as a centerpiece of a mix of transit-supportive land uses;

• Park and ride locations must be within the commuter shed for modal changes; and

12
 • Stations located partially or fully on or in private property should be considered (i.e.,
within developments or activity centers being served).

Stations are one of the components of light rail transit systems. Light rail connects
neighborhoods, districts and downtowns while stations are access points. Locating stations at
the right places with the right connections with other transportation modes is essential (David
Evans And Associates, Inc. and Sera Architects, Inc, 2009).

Generally, stations are selected places on a railway line where trains stop for:
• Exchange of goods and passengers;
• Control of train movements;
• Enable trains to cross from opposite directions;
• Sorting bogies;
• Changing engines and staff; and
• Taking diesel and water for locomotives.

Adequate distance between stations should be provided to maximize average operating speed.
Location of stations in close proximity to each other would be unreasonable for train
operations except in instances where stations provide essential distribution for passengers in
large urban areas. Closely located stations will reduce train operational speed. Hence,
minimum station spacing should be determined (Northern New England Intercity Rail
Initiatives, 2014). A review of standard practices in most North American cities indicate that
stations are generally located 400 m apart in the central business district, when acting as a local
service. In the outlying areas of the city, it is common for stations to be spaced at larger distances
apart, between 800 to 1000 m.

2.4.2 Types of stations

Categorization of stations is an important tool for use in planning and programming the size
and amenities of stations to meet local need, and in understanding the underlying factors that
determine stations‟ roles in the transportation system. Knowing station types are important
for planning station sites and for programming station sizes. There are three types of stations
based on their location relative to track; Online, Inline and Offline stations.

Amtrak‟s Station Program and Planning Guidelines (2012) have categorized stations into four
based on annual passenger flows, customer services and amenities consistent in station and

13
whether the station is staffed or un-staffed. According to Amtrak (2012), there are five
station categories according to the following bases:

Based on size
• Category 1 – Large;
• Category 2 – Medium;
• Category 3 – Caretaker; and
• Category 4 - Shelter.

Based on location

 Online station: Online stations are located within the vehicle runway and the transit
way so that vehicles can access them without leaving the runway. Online stations are
important elements of transit-way service speed, reliability, and accessibility.
 Inline station: Inline stations are located immediately adjacent to the vehicle runway,
typically along freeway interchange ramps. Although they require the vehicle to exit
the primary runway, they provide a fast access opportunity and immediately return to
the runway.
 Offline station: Offline stations support transit way accessibility and ridership, but
require transit way vehicles to exit the running way and require several turning
movements and potential traffic delays that impact transit way service speed and
reliability, especially during peak travel times. Figure 2.2 shows the layout of Online,
Inline and Offline stations.

Figure 2.2: Stations Based on Their Location Relative to Tracks (Online, Inline and Offline)

14
2.4.3 Station Location Planning/selection
There are many researchers who did studies on station location selection. (Scott A. Burger
and Bill Byme, 2015 ), explained factors they used to select rail transit stations and park-and-
ride location planning process as (land availability, highway access, proximity to major land-
uses, shared parking possibilities and transit-oriented development (TOD) that are potential
to influence decisions of system users. By taking these factors into account and reviewing
previous works, they proposed station location planning considerations.

Determining the location of transit stations is a critical part in designing a rail transit system
(Lai, 2012). There are five steps considered in locating rail transit station selection and these
are:
• Review the framework: understanding priority objectives and development strategy;
• Map target locations: determining potential station locations;
• Evaluate locations: refining target station locations through analysis;
• Consider the location type: determining location type and indentifying target sites ;
and
• Evaluation flowchart: refining target sites by using a set of defined criteria.

The above is very clear and convenient process, since it selects station location from
candidate station locations without knowing the alignment and decide alignment from
consecutive station, it cannot yield the real track alignment that satisfies various geometric
constraints (Lai, 2012). Another method of selecting station locations is placing along
selected alignment.

Station size and the complexity of its site design can vary significantly from location to
location; the site design issues included here are consistent across many station categories
and locations. The relationship of the station to the community, surrounding development and
other transportation modes is critical to its success.

Maps of population density, community centers, commercial districts or malls, other public
attracting places, amenities or institutions, park-and-ride locations and public parkades and
existing charging stations are very important inputs for mapping target potential locations
(LAB and TIPS, 2011).

15
Development of station area planning has three primary work areas: station location;
connectivity with other transport modes; urban design and infill development. Optimal station
location is a foundation for connectivity with other modes and for surrounding area
development (Development, June 2011). In order to attract many ridership, higher density
residential, higher density commercial, employment nodes and mixed use are potential land-
use to locate stations.

Locating, rail transit station is a key element in railway design. Station area planning should
be designed in which it can support people to easily access available transportation modes
since, Station locations have direct impact on passenger attraction, ridership, vehicle
operating speed and operational system such as travel time, riding comfort and cost of
operation (Lai, 2012).

2.4.4 Station location selection criteria

Proper site selection plays a vital role for both social and economic activities. It is obvious
that inappropriate site selection leads to heavy social and economic losses. Taking business
operations as an example, site selection is the first key factor and directly related to customer
groups, capital investment and recovery, and development strategy. Therefore, making good
preparations and analysis on the parameters of the site selection is absolutely necessary.

According to Jha et al. (2013), Selection of rail transit routes and intermediate stations depend on
many factors, such as demand, ridership, transfer stations, right-of-way, and construction costs.
The Metropolitan Council (2012) Transit way Guidelines identified four key transportation
connected site location factors based on primary or secondary factors for each mode: (i)
access to station for transit vehicles and customers; (ii) impacts on existing road network; (iii)
land availability for Park and Ride Plan; and (iv) railroad track-way operational impacts. The
others are land use- related factors which are:

• Land availability;
• Land types and costs (e.g., public right‐of‐way, joint‐use, private, etc.);
• Mix of land uses and compatibility with transportation functions;
• Development plans, including comprehensive and station‐area plans;
• Available infrastructure and the cost of providing additional infrastructures, including
bicycle and pedestrian infrastructure (e.g., sidewalks, bicycle‐pedestrian
overpass/underpass, etc.);

16
 • Proximity to affordable housing;
• Proximity to employment; and
• Size of and proximity to transit‐dependent, low‐income, and minority populations.

According to Metropolitan Council (2012), station site selection is based on:

• Acquisition of land;
• Proximity to town or village;
• Nature of land Area –gradient for station yards: 1 in 400;
• Approach roads to station site;
• Station site alignment;
• Site drainage –grade of 1 in 1,000;
• Station amenities;
• Type of station –Through type or Terminal type

Rail transit way stations should be sited to maximize convenience and minimize travel times
for subject passengers and vehicles under existing and planned future conditions. According
to David Evans and Associates Inc and SERA Architects Inc (2009), there are some ideal
characteristics of light rail transit station area considered in station locations such as good
connections, density, diversity and use. With increased interest in new transit systems for
congestion relief, optimal stations location planning is a principal activity.

Hence, there are various studies done on station site optimization based on different factors.
Optimization of station location is obtained by considering population map, employment and
transportation systems while site selection is done for the station to be easily connected to
other modes and to enhance surrounding area development (Development, June 2011).
Samanta et al., (2005:2-8) used a algorithm to optimize station locations along a rail transit
line so as to minimize the overall system cost (i.e., capital cost, operator cost, and user cost)
by using GIS and anti logarithm.

On related problem, Olowosegun Adebola and Okoko Enosko (2012), studied on

determination of best locations for bus stops. They used Slope Map, Shelter Map and Bus-
Stop Spacing Map to determine the best bus stop location. According to them, bus stops have
to be located in accordance to the following:
• Four hundred (400) meters interval on the major road;
• Available setback from the road ideal for shelter; and

17
 • Slope Map.
Railway station location selection is also a multi-criteria or factor problem. The factors
should be considered simultaneously as their effect is concurrent. Considering all criteria at
once is a complicated task. To handle such problems in transportation planning, GIS
combined with multi-criteria decision making software is best alternative.

2.4.5 Station accessibility evaluation

Station accessibility has two dimensions: (i) external accessibility (how easy it is to get to and
from the station); and (ii) internal accessibility that is ease of movement within the station
areas such as kiss-and-ride, park-and-ride, platform, concourse, etc. (MoTaC and ERC,
2009:9). The scope of this research is only external accessibility. This is because internal
accessibility to all users has been done previously.
Intermodal exchange and way finding
To ensure as seamless as possible journey, the station design has to take into account that
passengers do not start and end their journey in a station. Their travel experience is very
much determined by their approach and departure from the station. Depending on location
this can be any combination of the following forms of transport:
 Walking
 Cycling
 Car drop off or taxi
 Bus or tram
 Car parked in station car park
 Underground railway
 Airports

Interchange between all modes must be efficient and obvious to the traveler with clear
signage, minimum changes of level and where possible protected from the elements
(Network Rail, 2011).

18
 Figure 2.3: Station design checklist

A station is said to be accessible if it is easily accessed by modes such as pedestrians, buses,

private cars and bicycles. Pedestrians access station on foot walking a reasonable distance.
Many researchers have identified reasonable walking distances. Maximum acceptable walk
times at either origin or destination are quoted as standard maximum walk times five minutes
for access to bus stops and ten minutes for access to rail or LRT based on a walk speed of 80
meters per minute (4.8kph/3mph). In rural areas these times may be doubled.

Station accessibility assessment is done based on ideal station characteristics and

identification of missing elements for stations under consideration. Ideal characteristics, by
definition, are not always achievable and must be balanced with constraints and other
considerations. According to this report, ideal characteristics of stations are:

• Good connections
 Sidewalks connect the neighborhood to the station;
 Busy streets nearby have marked pedestrian crossings;
 Pedestrian crossings are provided over or across physical barriers (such as the
railroad or major busy streets) between the surrounding communities and the
station;

19
  Buildings are pedestrian-friendly (the front door or entrance faces the street or
sidewalk and there are ground floor windows);
 Streets and sidewalks are well lit;
 Bike lanes, multi-use paths, or low traffic streets provide bike access to the
station;
 Bus stops are located near the station, with clear paths from the stop to the light
rail station;
 Etc.

• Density, Diversity, and Use

 Station area includes a variety of housing types and densities such as apartments,
condominiums, townhouses/row houses, live-work units, and single-family
homes;
 Station area provides transit access to a variety of jobs and/or employment
centers;
 Retails, restaurants, and other commercial uses present an opportunity to be
supported by transit;
 Station areas include institutions such as schools, parks, and medical facilities that
would benefit from transit service;
 Etc.

• Opportunities for new uses

 There is vacant or underutilized land near stations;
 A light rail station could support or encourage new development consistent with
the city and neighborhood plans and policies;
 Etc.

• Get the land uses right

 Make retail strategy market-driven, not transit-driven;
 Develop mixed-income housing and encourage every price point to live around
transit;
 Segregate uses and vary density where appropriate. Mixed uses don‟t have to be in
the same building;
 Allow employment areas near the station to promote reverse commuting and
optimize transit capacity.

20
 • Build a place, not a project; ensure good urban design
 Design the station as a positive edge, center of, gateway to, or main amenity for
the station community;
 Use high quality urban form to support mixed incomes and uses;
 Balance higher densities with additional amenities (e.g., attractive design,
neighborhood-serving retail and services, functional open spaces);
 Create landmarks and beacons;
 Preserve and invest in existing neighborhoods;
 Extend the interconnected street grid and establish a finer-grained lot and block
structure in new or redeveloping neighborhoods;
 Transitions into the surrounding single-family residential areas of a neighborhood;

• Get the parking right

 For station locations that must accommodate park-and-ride facilities, locate them
within a five-minute walk of the platform but not directly in front of the station;
locate utility structures so as not to preclude redevelopment of prime, station
proximate sites.
 Discourage station spillover parking into the neighborhood: Unless a park-and-
ride facility is provided, stations are meant and designed to serve the surrounding
neighborhoods and land uses and be accessed by foot, bike, or bus rather than by
automobile. On-street parking should be managed to avoid spillover from the
station through programs such as parking permits and metered parking.

By using these checklists, any station assessments can be done. In the case of this research,
only external accessibility evaluation of station location has been done.

In the process of creating active urban transportation, railway transit system as a means of
mass transport, should be accessible to all modes and pedestrians and stations should be
easily and naturally accessible, with a clear layout for street-to-train, bus-to-train, and station-
to-area destinations.

A primary factor in locating a single station for all transit-way modes is creating access to the
station for transit vehicles and customers. It is critical to ensure that customers and transit
vehicles, including those specific to the mode and those connecting for transfers, have safe
and convenient access to the station. Similarly, convenient access intern will ensure efficient

21
transit operations for all modes, and is critical in providing fast and reliable service on the
transit way.

If a station is located at a place where other modes of transport are accessed or linked to, the
station is defined as intermodal connected station (Northern New England Intercity Rail
Initiatives, 2014).

An intermodal connected station attracts many riderships, since families could choose to own
one automobile instead of two because they will be confident that other transportation modes
will enable them to move conveniently throughout the city. Shifts in transportation modes
will yield a significant benefit to personal and urban health and to environmental
sustainability

The proximity of stations to rapid transit, intercity and commuter rail, local and intercity bus,
major roadways, or airports will improve intermodal connection between travel modes and
attract maximum rider‟s mobility (Northern New England Intercity Rail Initiatives, 2014).
Passenger access to railway station can be through private automobile, taxi, transit, bicycle
and walking. Therefore, proper station location planning shall guarantee that each mode is
accommodated and transfers are provided.

The Guideline described types of station access that should be considered when locating a
station and selecting the station type as transit way, connecting roadways that support transit
transfers or customer access, sidewalks, and trails.

Railway transit system accessibility is analyzed based on station accessibility analysis; and
station accessibility will be analyzed based on the following station access criteria:

• Private automobiles and taxi pick-up facilities areas for private automobile, group
ride, and taxi stands should be provided to facilitate easy passenger drop off and exit
from stations. Easily identifiable taxi stands should be built in close proximity to the
main station entrance;

• Parking availability for park-and-ride and kiss-and-ride;

• Private car rentals and car sharing;
• Transit connection: Bus stops should be located near LRT stations so that passengers
getting off the bus and walk the shortest possible distance to access the station;
• Pedestrian network/ walkways; and

22
 • Bike and bike sharing.
The average straight line distance of each access mode may differ from each other.
Comparing the walk and private car mode, a large number of people who walk to the station
have their point of origin within a 500-meter radius of the light rail transit station. While for
the private car mode, most of the origins are farther from the center. The mean of the straight
line distance for walk mode is 500m, and for the private car mode is 1500m.

Light rail transit (LRT) line may be constructed having its own right-of-way or in the
medians of road /elevate or at grade. For many advantages, LRTs are constructed in road
medians worldwide. Some of advantages of such stations include avoiding high right-of-way
acquisition costs, increasing train speeds, and limiting modal conflicts.
In Los Angeles LRT networks, they used seven criteria and reviewed pertinent literature to
analyze accessibility of elevated stations. These criteria were: 1) cumulative opportunities; 2)
cumulative dis-amenities; 3) infrastructural barriers; 4) connectivity; 5) station visibility and
legibility; 6) platform accessibility; and 7) environmental conditions.

2.5 LRT Station location planning using AHP and GIS

2.5.1 Analytical Hierarchy Process (AHP)
AHP is a multi-criteria decision making tool, which was proposed by Saaty (1980) and is a
structured technique for decision making based on a hierarchical framework constructed
through mathematical pair-wise comparisons.
The weights for the decision making criteria are derived based on the pair-wise comparisons
of the relative importance determined by experts between each two criteria (the sum of the
weights equals to 1). As suggested by Saaty (1994), a judgment or comparison is the
numerical representation of a relationship between two elements that share a common parent.

AHP is the most creative multi-criteria decision making technique. It structures evaluation
factors descending from an overall goal to criteria, sub-criteria and alternatives in successive
levels (Saaty, 1990).

AHP was framed as a multi-criteria procedure and integrated with Geographic Information
System (GIS) to develop an optimal transit station site assessment system and accessibility
analysis with the flexibility to account for the population in the study area, land use as well as
proximity to parking buildings and bus stations/stops and distance from road network and
LRT lines.
23
And GIS is a powerful tool to develop better public transportation analysis technique due to
its ability to manage, recall and evaluate information effectively. GIS has many applications
in transport planning, railway operation management, resource allocation, optimization of
railway alignment and station locations and in handling and analysis of different spatial data
related to railway within a single geodataset.

Once all spatial data within study area have been collected into a geolocated dataset, the
information can be used to inform the engineering works are required and so on. For
example, any locations where railway alignment deviates or failed from existing track
alignment can be quickly identified and resource needed for track repairing can be allocated
to the place. The list below identifies the major functions or disciplines in which GIS has
been successfully deployed in railway organizations:

• Real estate management; • Environmental and construction

• Facility management; management;
 Track,
• Intermodal management;
 Power, and
• Passenger information;
 Communications and signaling.
• Capacity planning;
• Asset tracking; • Marketing;
• Commodity flow analysis; • Supply chain management;
• Emergency response management; • Site selection; and
• Risk management.

Since AHP is a comprehensive approach due to its ability to consider various simultaneous
criteria in the decision-making process, this paper will highlight on the application of multi-
criteria techniques incorporated with GIS in identifying possible LRT station that take into
consideration important and critical spatial criteria. Only through the use of GIS is it possible
to visualize all of this information simultaneously and quickly, and thus easily understand all
the relative impacts that can result from the station under consideration.

Applying AHP to a decision-making problem involves the following four fundamental steps
(Piamviriyawong, 2006).
• Model specification: First, feasible investment alternatives are specified followed by
the determination of criteria for the evaluation of alternatives. These criteria are
further grouped into logical categories.
24
 • Pair-wise comparison of categories and criteria: The relative importance of criteria
within each category and of each category within the group of categories is
established through pair-wise comparisons using a square matrix structure. The values
of importance are taken from Saaty‟s 1-9 scale mentioned above.

• Weighting of investment alternatives: Every investment alternative is rated with

respect to every investment criterion in the evaluation model. Pair-wise comparison is
applied to obtain weighting for qualitative data. If quantitative data are available, then
the weighting is done by using the existing or estimated performance data.

• Investment rankings: Finally, weighting of the alternative is combined with the

weighting of the criteria to form an overall rating for each investment alternative. The
alternative with the highest weight is ranked as the best choice, taking into account
the relative importance of each criterion and the relative desirability of the alternatives
with respect to each criterion.

(Saaty, 2008) also clearly confirmed the following AHP processes that should be undertaken:
• Identifying the problem by identifying the goal;
• Identifying the criteria that affect and influence the problem;
• Identifying the suggested alternatives or solutions that will be compared and
differentiated;
• Constructing the hierarchical model, including the higher level that represents the
desired goal; the middle level that represents criteria; sub-criteria that influence the
problem; and the lower level that represents the suggested alternatives and solutions
that will be compared and differentiated in order to solve the problem;

• Collecting data, noting that it is required here to identify the personal judgments of
the decision makers, the experienced people and researchers who are familiar with the
problem in order to provide solutions for it. This is accomplished by conducting
personal interviews or through special questionnaires related to the analytical
hierarchy;
• Designing pair-wise comparisons matrix, where elements are compared by two-way
method for each trait; and
• The pair-wise comparison starts from the top of the hierarchical model, which
represent the goal, as mentioned before, the highest level in the model (the goal). The
beginning is done by comparing the criteria relative to the goal where this comparison

25
 follows the pair-wise method by comparing between two criteria for the same goal
followed by making a shift toward another two criteria to select the goal.

These are steps used in four levels AHP model. While three level AHP model is similar to
four levels but there is no fourth level in three levels AHP model. In this research three level
AHP model was used.

2.5.2 Geographic information system (GIS)

Geographic Information System (GIS) is a system designed to capture, store, manipulate,
analyze, manage and present all types of geographical data that is spatially geo-referenced at
all scales. In the simplest terms, GIS is the merging of cartography, statistical analysis and
database technology. The GIS applications are tools that allow users to create interactive
queries, analyze spatial information, edit maps and present the results of all these operations
(Clarke, 1986).

2.5.3 Application of GIS in Railway Engineering

GIS applications cover much of the broad scope of transportation. Transportation analysts
and decision makers use GIS tools in infrastructure planning, design and management, public
transit planning and operations, traffic analysis and control, transportation safety analysis,
environmental impacts assessment, hazards mitigation, configuring and managing complex
logistics systems, just to name a few application domains. Similarly, GIS is a useful tool that
can be applied in many decision making problems. It helps in decision making process by
handling and considering simultaneous factors/ criteria that affect decision making.
Rail way engineering needs GIS in different decision making problems such as:
• Data handling in a single geo-dataset; many railways related data with different
attributes that can be stored, edited and analyzed for planning and design purposes;
• Resource allocation;
• Railway system planning;
• Identification of areas that have drainage problems and flood area delineation;
• Optimal railway alignment and station site selection;
• Station access evaluation;
• Earth work estimation (cut and fill);
• Railway system operation management;
• Railway maintenance management; etc.

26
Most GIS have very limited capabilities for integrating geographical information and the
decision maker‟s preferences. Therefore, it is suggested there should be integration between
GIS and AHP.

This research is envisaged to identify optimal location of urban railway transit stations which
helps to create more convenient transportation options by reducing congestion, primarily
based on population density map, distance from collector road network, proximity to parking
building and bus/taxi stops/stations map and land use and site slope condition maps taking the
Addis Ababa LRT. In addition, accessibility of phase one LRT stations has been studied in
this research.

27
 CHAPTER 3
3 METHODOLOGY
3.1 Introduction
This study adopted an analytical, descriptive method by using the analytic hierarchy process
(AHP) integration with geographic information system (GIS). This was done first, through
station accessibility evaluation and site suitability for new stations along with LRT lines; and
second, by developing or through identification of criteria were the major factors considered
in the methodology. In the case of new station site suitability analysis, development of LRT
station site selection decision determinant criteria was performed followed by weight
determining process criteria using AHP Super Decision software.

The first step to achieve this goal was the development of questionnaires where 8 experts
were asked to determine the relative importance of each factor. The method evaluates the
relative significance of all the parameters by assigning weight for each of them in the
hierarchical order; and, at the last level of the hierarchy, the suitability weight for each class
of the factors was given. Typically, the priority of each factor involved in the AHP in order to
produce the land suitability map, the actual factors and class weights (ratings) of the
parameters involved in the study are needed. These are determined systematically based on
the AHP analysis.

The relative importance of the criteria was judged by experts who could evaluate the relative
significance of the criteria that determine station location decision. Each respondent was
asked to compare all of those factors and to rank them based on her/his preference. To help
the respondent in evaluating those factors, the Analytic Hierarchy Process (AHP) program
was used. AHP allows respondents to judge those key factors using a pair wise comparison
method. With this approach, the respondent compared two factors at a time, decided which
one is the more important factor in determining locations of transit alignments, and set the
preference level of the selected factor. The comparisons were done using actual
measurements or based on the relative subjective evaluations of the respondents.

By using relative importance matrix judged by experts, weight criteria were determined using
AHP excel template. Then, the weights were used as an input in GIS multi-criteria spatial
analysis for suitable station location identification.

28
Station site selection criteria were determined by reviewing related literatures. Criteria
weights were determined by means of AHP which were then used as inputs in weighted
overlay of these criteria in ArcGIS for identifying suitable station sites.

3.2 Methods Used in New Station Accessibility Analysis

The following figure 3.1 shows methods flow chart used in station accessibility evaluation.

Figure 3.1: Methods Flow Chart Used in Station Access Evaluation

3.3 Methods Used in Station Suitable site Selection

As explained above, criteria development, relative importance of criterion judgment by
experts, criteria weight determination using AHP/ANP and GIS Spatial Analyst tools were
applied in potential station site analysis. Figure 3.2 shows methods flow chart in suitable
station site analysis.

29
 Map Rasterization

Figure 3.2: Methods Flow Chart in Suitable Station Site Analysis

Lists of activities in this topic are:

3.3.1 Euclidean Distance Calculation

The Euclidean distance operation describes each cells relationship to a set of sources. The
function gives the distance from each cell center to the closest set in the raster format data set.
This function was used to calculate distances that are distance from bus station and parking
building to the stations. The Euclidean distance output is a raster file format that has different
layers or concentric rings from the point of reference. The distance is given in linear units
meters, kilometers, miles etc. In this study case, Euclidean distance calculation was done for
distance from road and from LRT lines.

3.3.2 Buffering
A buffer is an area formed from a locus of constant radius from a given point or along a
linear feature. This analysis can be done on either points or lines and the result is always a

30
new polygon layer. These buffer polygons define the inside and outside of a region which are
typically distinguished by different codes in the attribute table. In this project, buffering was
done on all the bus stops/stations and also on all Multi-storey parking buildings aimed at
identifying proximate bus stops and parking building to LRT stations.

3.3.3 Rasterisation
Rasterisation (or vectorization) is the task of taking an image described in vector graphics
format (shapes) and converting it into a raster image (pixels or dots) for output on a video
display or printer, or for storage in a bitmap file format. And Rasterisation refers to
conversion from vector to raster data.

3.3.4 Reclassification
This is the process of assigning weights to different classes of a raster data set. This process is
aimed at showing how the different concentric layers are placed and how suitable are they. In
this project the datasets reclassified were slope, Euclidean distance of railway road, Euclidean
of LRT line, land use in raster, and population in raster format too. Values were assigned to
each layer on a uniform scale of 1-5. This scale is thus referred to as the suitability scale.
Since all the data was harmonized by this process, the suitability process was possible to be
carried out.

31
 CHAPTER 4
4 DATA COLLECTION
4.1 Study Area
Addis Ababa is the capital city of Ethiopia and the place for the Head Quarter of the African
Union and it is sometimes referred to as “the capital city of Africa”, due to the presence of
different international organizations. Addis Ababa is attracting conference attending tourists
as a result of its political location and booming economy.

According to the 2007 census data, the population of Addis Ababa was 3 million (Central
Statistical Authority 2007 Report). The area of the City is about 529 km2; organized into 10
sub-cities and 99 woredas. Due to its location at the center of Ethiopia, Addis Ababa is a
transport hub connected to different parts of the country.

As a result of fast economic growth in the country, Addis Ababa is transforming from a
predominantly administrative and service center into an industrial and financial center which
intern leads to high passenger and goods mobility.

Currently, Addis Ababa‟s transportation system has various problems: aged fleet, chaotic
movement of mini-bus taxis, unacceptable carbon emissions which is hazardous to life and
property and unsafe and unreliable (Yehualaeshet Jemere, 2012). More than seventy per cent
of the vehicles in the Country are operating in Addis Ababa with three million populations
whereas, thirty percent of the vehicles are operating where more than ninety-five of the
population are settled which is not reasonable.

Bus and taxi comprise Addis Ababa‟s public transport. However, walking is the predominant
transport mode. So far, Addis Ababa has solely relied on traditional buses for passenger mass
transit. However, currently though, the construction of light rail transit (LRT) is well
underway. Since, the center of city is highly populated and has high traffic congestion, transit
services beyond traditional buses is urgently needed and was the reason for the construction
of LRT. In this regard, there are two lines under Phase I of AALRT; East-West (E-W) and
North-South (N-S) lines. The E-W line has a total length of 17.4 km and the N-S line has
total length of 16.97km. The numbers of stations along both lines are 39 of which five are
common to both lines.

According to information in conceptual design document, there are twenty-two LRT stations
placed under Phase I of E-W route project, five of which are shared with N-S route. The

32
average interval between the two adjacent stations is 815 meters. The longest interval is 1,210
meters and the shortest interval is 525 meters. Similarly, twenty-two LRT stations are placed
along N-S route, five of which are shared with E-W route. The average interval between the
two adjacent stations is 793 meters. The longest interval is 1,370 meters and the shortest
interval is 510 meters.

The study area for the stations' accessibility analysis of this research is the whole Phase I
AALRT networks. While, the whole phase two or phase one line extensions are for suitable
or optimal station location selection.

Figure 4.1 below shows the AALRT Phase I network and its future extension. Similarly,
figure 4.2 shows the AALRT network and its future extensions. Both were provided in
Bankable Feasibility Study for Addis Ababa Light Rail Transit, 1st Draft, and Volume I,
2009.

Figure 4.1: Addis Ababa existing and future extension LRT network

33
Figure 4.2: Study Area for existing stations accessibility analysis

34
4.2 Primary Data
4.2.1 General
The primary data were collected from experts through questionnaires. The experts were
selected based on their knowledge on transportation planning. Depending on AHP model; a
questionnaire was developed which was presented to a group of specialists whose members
were anonymous to one another. The process elicits qualitative judgments that indicate the
strength of a group of decision makers‟ preference in a specific comparison, according to
Saaty‟s 9 point scale (Shahroodi et al, 2012). If the members of the group know one another,
the decision of one could be influenced by the decision of the other leading to poor
judgments. In view of the above, data were collected. In the case of discrepancies arose
between judgments of the experts, the researcher would provide more explanations to those
concerned for more thoughts in reaching consensus.

4.2.2 Participants
The study was applied to a group of 8 experts that encompassed different specialists and
students working in transport planning related. They are selected based on their expertise and
experience. The data collection was administered on those specialists in order to develop the
weights.

AHP is a subjective method that is not necessary to involve a large sample, and it is useful for
research focusing on a specific issue where a large sample is not mandatory (Cheng, 2002).
Cheng and Li (2002) pointed out that AHP method may be impractical for a survey with a
large sample size as „cold-called‟ respondents may have a great tendency to provide arbitrary
answers, resulting in a very high degree of inconsistency. AHP survey with a small sample
size has been conducted in previous research. For example, Cheng and Li (2002) invited 9
construction experts to undertake a survey to test comparability of critical success factors for
construction partnering.

4.3 Secondary Data

Secondary data is normally collected through examination of existing documents in pertinent
organizations. In this regard, secondary data were collected by gathering raw data (shape
files, AutoCAD files, population and travel demand). Table 4.1 shows the list of required
data. Collected data were converted to format that is used in GIS dataset. Table 4.1 shows the

35
 required data and table 4.1 shows the Addis Ababa road network and population density as
provided by AACRA.
Table 4.1: Required Data

Source Data Format

Data List Organization Required Data Format
Road network(GIS shape file) AACRA Shape file AutoCAD
Land-use AASOIDPO Shape file Shape file
Ortho-photo AASOIDPO Raster Raster
GIS shape files( phase I & II stations) AA LRT office Shape file

Population (point Data) and socio- Federal central Population Raw data
economic data (population density) statistics agency density in
raster map
AA DEM(Digital Elevation Model AASOIDPO
GIS software (ArcGIS 10.2)
AHP( Super Decision) software

Figure 10: Addis

Ababa Road
Network from
AACRA Converted
to GIS shape file and
Map 42 AA
Population Density

Figure 4.3: Addis Ababa Street Network Converted to GIS shape file and Population Density

36
 CHAPTER 5
5 PHASE ONE STATION ACCESSIBILITY ANALYSIS
5.1 Station Access Guidelines and Standards
This chapter proposes a practical rail transit station external accessibility assessment
methodology based on guidelines and standards acquired from literature review. Since
passengers do not start and end their journeys at stations, the station design has to take into
account that their travel experiences are very much determined by their arrivals and
departures from stations.

Depending on location, this can be any combination of the following forms of mobility:
• Walking;
• Cycling;
• Car drop off or taxi;
• Bus or tram;
• Car park in station;
• Underground railway; and
• Airports.
Hence, station accessibility both external and internal should be considered during station
design. The scope of this study is only external accessibility which considers station
intermodal connection because; internal accessibility assessment has studied previously.
Furthermore, station accessibility should be evaluated based on guidelines and standards. The
guidelines and standards that followed by access mode are intended to support
implementation of the access hierarchy with the goal of achieving an optimal balance of
access to the transit system.

• External Connectivity

Connectivity implies to connection the station has with other available transportation modes.
An intermodal transit station (ITS) can provide the foundation for greater integration of
future transportation services, increased connectivity to other regional transportation
providers, reduced rates of vehicle-miles traveled (VMT), and reduced congestion. In
addition, development of an ITS can leverage investment that supports economic
development, environmental benefits through reduced energy consumption and air pollution
emissions, and enhanced quality of life .

37
Accessibility of station can be enhanced through the design of compact urban environments
and interconnect networks of streets at neighborhood and urban scale. And this will be
obtained by integrating transport plan and the city‟s land use plan (Tesfaye, 2012).

According to David Evans Associate Inc. and Sera Architects (2009), the following checklists
are characteristics of Ideal station location that can be used as a reference to check missed
elements of stations. These ideal characteristics may not be fulfilled but, help to assess
accessibility of stations relative to absolute station.

5.1.1 Pedestrian Access Guidelines and Standards

According to Regional Transportation District (2009), the following walk speeds shall be
used at station facilities for transit access:
• Level walk speed: 200 feet per minute
• Horizontal component of stair walk speed: 130 feet per minute
• Appropriate adjustments for barriers, such as an elevator or street crossing, shall
be added to the walk time.

At stations where elderly or disabled persons comprise an estimated 20% or more of

transfers, a horizontal walk speed of 150 feet/minute and the elevator time shall be used
instead of stair walk speed.

For distance calculation purposes, the centroid of the platform shall be taken from the
distance between the front of the trains (highblock) in each direction and the extreme edges
of all the platforms. The bus connection points are the designated gates.

5.1.2 Bus Transfer Access Guidelines and Standards

For rail-bus connections, knowledgeable commuters generally have above average walk
speeds, but less frequent and new riders require more time to find their way. Crowding at
stations during peak periods also impedes free-flow, and elderly and disabled passengers
need more time to make their connections. Sight distances also affect perceived convenience.

The proximity of the bus gates to the train platforms is governed by establishing the distances
(and travel times) that provide for desirable, average and maximum transfer connections. The
probability of making a connection is also of concern. A missed connection means the
passenger will wait a time equivalent to the headway (time between buses) on the connecting

38
route. Table 5.1 shows rail-bus transfer distance standards as provided in Transit Access
Guidelines, Version 2.4 (2009).
Table 5.1: shows Rail-Bus Transfer Distance Standards

Transfer Time Maximum Walk Maximum Arc Weekday Bus

↓ Path Distance Distance Trips
Desirable < 2 minutes 125m 75m >75%

Average = 3 ½ minutes 210m 130m >85%

Maximum < 5 minutes 300m 180m 100%

5.1.3 Bicycle Access Guidelines and Standards

All stations, regardless of whether they have park-n-ride facilities, shall also have bike
parking. Parking shall be provided in the form of bike racks, bike lockers, bike station, or a
combination thereof.

5.1.4 Auto Access

These guidelines and standards cover both kiss-n-ride drop-off/pick-up areas as well as park-
n-Ride facilities. It is important to recognize that in certain instances, competing demand for
space or physical site constraints may make the guidelines impractical. The kiss-n-ride
facilities at stations typically include areas used for dropping-off and picking-up transit
passengers, as well as taxi stands, provisions for paratransit vehicles and private shuttle
buses. It may be possible to combine kiss-n-Ride and transit areas provided that automobiles
not delay transit vehicles (Regional Transportation District, 2009).

5.1.5 Kiss-and-Ride

Guidelines

These facilities should be convenient for both pedestrians and motorists to use, or else they
will find other locations to engage in pick-up/drop-off activity that may cause desirable
conflicts. The kiss-and-ride should be designed to maximize vehicle turnover, facilitate
traffic flow, and avoid conflicts between pedestrians and other access modes and vehicles.
One-way traffic flow is recommended and the site should allow for re-circulation.

39
Standards

All stations, regardless of whether they have park-n-Ride facilities, shall also have kiss-n-
Rides when practicable, that are sized to meet forecast or demonstrated demand. Stations
located in TOD areas may be accommodated by on-street kiss-n-Ride facilities, subject to the
review of local jurisdictions. Except where prevented by physical site constrains, the kiss-n-
Ride shall not exceed a walk distance of 400 feet from the platform center, and a maximum
arc distance of 240 feet. The kiss-n-Ride shall have a direct line of sight to the station
entrance. Pedestrian crossings from the drop-off/pick-up lane shall include a stop sign and
marked crosswalk. Signage shall direct both vehicles and passengers existing stations to
drop-off/pick-up areas. Figure 5.1 below shows all facilities or elements needed at LRT
stations as provided in Accessing Transit: Design Handbook for Florida Bus Passenger
Facilities, Vol. III, 2013

Figure 5.1: Typical Intermodal Railway Station Interface with Bus

40
5.2 Results and Discussions
5.2.1 Criteria Mapping and Analysis
This chapter evaluated the accessibility of constructed phase one station sites for pedestrians,
bicycles, intermodal transit, and (At the moment, kiss-and-ride is not common in Ethiopia)
passenger-drop-off vehicles. Based on these criteria, access of the stations evaluated and
some recommendations have been done. Passenger flows within light rail transit system
redistributed only at stops and boarding / alighting to the street network (China Railway
Group. Lt., 2009). For LRT station to meet this target, station site should be integrated with
other transportation mode such as integration / connection with collector road, pedestrian and
bicycle walkways.

5.2.1.1 Current Addis Ababa LRT system characteristics

As mentioned above, there are twenty-two LRT stations under Phase I of East-West route
project; five of which are shared with North-South route. Average interval between two
adjacent stations is 815 meters. The longest interval is 1,210 meters and the shortest 525
meters. In view of the above, accessibility of AALRT stations were analyzed based on the
following station design criteria identified after intensive literature review. The following
factors were considered in evaluating accessibility of Phase I Addis Ababa LRT stations.

• Inter Connectivity
 Road network;
 Accessibility to Pedestrian;
 Accessibility to Bicycle users to station;
 Park-n- Ride availability; and
 Bus and taxi transfer station.

• Ridership
 Population density

5.2.1.2 Connectivity

(i.) Road Network

The line for LRT should be coordinated with the existing roads and planned roads in the
future. One of the most significant hard factors in influencing walk trips is the presence of a
cohesive and efficient street network. The key to providing pedestrian access is to reinforce

41
the principle that streets should be designed for all travel modes, not just cars. In the figure
5.2 below, road network of the study area is used for connectivity analysis within 800m
radius surrounding LRT stations.

Figure 5.2: Study area LRT station connectivity with other modes

42
(ii.) Park-and-Ride

Park-and-ride is the act of parking at a customary built surface car parking area, on the city
outlets and transferring to public transport to travel onward to one‟s destination. The main
objective of building light rail transit in Addis Ababa is to meet the current large transport
demand and traffic congestion. Since the AALRT system has large capacity to move mass
passenger at a time, it can meet the demand. At the same time, it has to solve congestion
problem. It should be noted that the LRT infrastructure will also induce problems because it
replaced road lanes in some areas and be obstacle in places of at-grade crossings by
increasing car turning time.

In view of the above, for AALRT to reduce traffic congestion in addition to transporting
mass, system stations has to provide parking or shall be close to public parking buildings.
Stations at suburbs should have parking areas because, in sub-urban locations, there are few
origins and destinations within close walking distance and auto travel is predominant mode,
and there is more space for large parking lots. But intermediate stations can optionally
combine parking. Similarly, parking facilities should be provided at suburban area of the
LRT stations. So, parking should be provided at some stations, depending on modes of access
used by the riders to be served in the area. This suggests community in service area of
stations of 400m-800m radius away from geographic center of stations may access LRT
stations by various modes such as walk, cars, bicycle and bus.

Densely populated urban areas or sub-urban centers have a significant fraction of transit
riders that access transit services on foot in order to avoid the costs associated with
owning/driving/parking a vehicle, while suburban areas typically have many transit riders
relying on access by auto. Denser projected land-use assumptions near transit lines, which
resulted in significantly higher ridership projection (Byrne, 2008)

Walking-based stations should be easily accessible to pedestrian while, park-and-ride should

have easy access to the existing road network and preferably be near the roadways carrying
significant traffic volumes. Personal vehicles need to be accommodated at stations to
encourage transit-way ridership, which is usually done through a park-and-ride lot where the
customer parks his/her personal vehicle and rides transit to their destination.

Consequently, light rail transit stations should be sited to maximize convenience and
minimize travel times for transit-way passengers and vehicles under existing and planned

43
future conditions, with primary emphasis on existing conditions. This study proposed that
downtown stations can take advantage of existing surface parking and proposed multi-storey
parking buildings while suburb stations can use park-and-ride under the proposed City Master
Plan. Additional park-and-rides may not be needed at the suburb stations since the transport
planning team at AASOID have considered current and 10 to 25 years projected park-and-
ride demand of the areas. Figure 5.3 depicts the proposed city center parking building
multiple-ring buffering in Addis Ababa.

Figure5.3: Proposed City Center Parking Building Multiple-Ring Buffering

Figure 5.3 shows multiple ring buffers with different radius of parking building proposed to
be constructed in Addis Ababa is overlayed by Phase I LRT stations. If parking buildings are
available in walking distances to LRT stations, passengers can park their private cars and take
mass transit to their destinations contributing to decrease in current traffic jam.

Observing Figure 5.3 most of the stations are within walking distances (400m) of and some
of them are within 800m from one or more parking buildings. Even if provision of parking

44
 buildings at intermediate stations is optional, the above stations which are within walking
distances from parking buildings are almost intermediate stations.

Suburb stations should provide parking for park –and-ride users that access stations by
driving. Park and ride stations are more common on the outskirts of the downtown core so
that commuters can park near the station and access the downtown via LRT (CMR, 2009).

Figure 5.4 shows the rough map of park-and-ride site proposed in the Addis Ababa City
Master Plan as provided in the AASOID Transport and Road Network Planning Report, May,
2015

Figure 5.4: Park and Ride Locations in Addis Ababa

Park-and-ride will attract riders who would not take transit with either BRT or LRT transfer.
Locations of the facilities are: Lebu, Ayat, Gelan, Akaki, Addisu Gebeya, Asko and Tatek.
These park-and-rides are located at strategic points throughout the city transportation network
to attract and encourage people to integrate public transit into their trip decisions.

(iii.) Pedestrian and Bicycle Access

Pedestrian and bicycle traffics are major sources of transit ridership. In reality, whatever their
original access mode, all transit patrons are converted to pedestrians before boarding and
immediately after alighting. Pedestrian access, therefore, is the highest priority access mode,
and transit facilities should reflect this in their design and layout.

45
All transit trips begin and end with a pedestrian trip regardless of the travel mode between
origin or destination and the transit system. For that reason, the conditions that support transit
access have the additional benefit of supporting pedestrian and bicycle trips for shorter
distances by creating more choices for travel, high quality transit access maximizes the
benefits of the infrastructure investment in transit both capital and operating.

Pedestrian routes or sidewalks are how all transit riders eventually get to and from the station
that provide pedestrian access to the neighborhoods and services surrounding the station. A
well-developed sidewalk network, with high quality pedestrian treatments, that connects the
station to the neighborhood and nearby services is vital to providing effective and efficient
station access from as many routes as possible. Like vehicle facilities, sidewalk corridors
should mirror the urban context for example, should be wider along main streets, or have
sidewalks and tree lawns within single-family neighborhoods and be wide enough for
expected foot traffic, particularly around train stations.

Similarly, pedestrian routes that arrive at the station should continue past the station property
edge to the platform entrance. To serve all riders, sidewalks need to be designed based on the
physical needs of wheelchair users and visually impaired in order to have consistent,
equitable, and reliable access. This access would include minimum 1.5-m-wide sidewalks
with seamless transitions between streets and other surfaces. Research shows that people are
willing to walk about a 400-meter, or 10 minutes, to light rail stations. The sidewalk
improvements and network should reflect this distance along the actual street network to
station entrances.

A one km pedestrian access distance is normal in suburban conditions and within 500m in
urban conditions are favorable thresholds for a higher capacity rail transit system. The draft
stated that in the more densely populated areas this radius should be constant, as operations
may be affected if the stations are too close together. In general this distance provides an
optimum rapid transit system performance whilst maintaining a reasonable walking distance
to any station along the transit corridor.

In 2006, the Addis Ababa modal share was pedestrian 45% and bus transport was about 46%.
Private car modal share is only 9%. This implies that pedestrian and public transportation are
taking the lion‟s share of Addis Ababa transport systems. So, this needs good intermodal

46
transit planning. Figure 6.5 shows the Addis Ababa modal share by 2010 as provided by
AASOID Transport and Road Network Planning report May, 2015.

Modal Share 2010

5%
Type of Mode
24%

55% Walking
Buses
Taxis
16% Cars and other

Figure 5.5: 2010 Addis Ababa modal shares

For AALRT system to meet the intended aim stations should be accessible to pedestrian, bus
and bicycle traffics since most passengers are pedestrians. New investment in urban transit
tends to be oriented to one of two accessibility options; either convenient walk-up
accessibility or the provision of significant park--and--ride. The type of access depends on
how and where transit planners locate transit stations in the urban environment.

According to new Addis Ababa master plan road network, there are some non-motorized
lanes dedicated to pedestrians and bicycles and there are pedestrian sidewalks along many
collector roads. With existing pedestrian-walk--way network, it enables pedestrians to access
motorized public transportation mode LRT and BRT after walking reasonable walking
distances which in turn encourages healthy community. Figure 5.6 shows circular station the
service area or catchments of 500m radius.

47
 Figure 5.6: Circular station service area within catchments of 500m radius
The figure above shows that station catchments overlap, meaning people within area of
overlapping can use one of near stations.

(vii.) Bus/Taxi transfer stops

To substantiate the high cost related with the built and operation of rail transit system,
sufficient trips need to be attracted from alternative modes so as to generate revenue and
reduce congestion. Hence, bus access should be provided at each station for various local,
cross-town, express, and circulator/shuttle routes, depending on the station. Bus/taxi stops
should be added for all stations and would typically be adjacent to the rail platform to provide
convenient transfers for passengers of long distance because LRT does not provide door to
door services.

Bus stops are located near the station, with clear paths from the stop to the light rail station.
According to these draft ensuring, efficient transit connections to and from rail stations is

48
crucial in the development of a seamless, well-integrated transit network. Transit stops and/or
transfer way finding signage should be immediately visible upon exiting the platform. Well-
located bus stops will minimize walking distances to and from the platform entrance and
avoid the need to cross roadways, particularly busy arterials. Minimizing distances between
bus stops facilitates bus-bus transfers and simplifies bus-light rail transfers. In some station
areas that have pedestrian-friendly walking environments and where crossing busy streets is
not necessary, short walks between the station platform and alternative modes of
transportation may help support street-level retail and an active station area.

Bus stops should be located on-street in travel lanes with curb extensions (unless off-street
facilities are necessary to accommodate layovers or transfers) to prioritize bus movements.
Bus stops should not be located where they will block crosswalks, obstruct traffic signals, or
be obscured from motorists, bicyclists, and pedestrians.

The study area stations case, there are bus stops within 500m catchment. This implies bus and
taxi riders from outside of the catchment area can access LRT stations. Figure 5.8 below
shows access modes.

49
Figure 5.7: Connectivity to Bus top, park-n-ride and taxi stops

50
 As explained above, parking should be provided at Saris, Abbo and Kaliti stations and
optionally at downtown or city densely populated area as people prefer walking to driving to
stations because of its minimum distance from homes. Provision of parking can reduce
congestion on this direction at some points because car riders may use LRT during peak hour.
Regarding bus transit transfer access, almost all stations are within short distance from bus
stops/stations. As a recommendation in this case, multi-storey parking or surface parking
should be provided at suburbs stations of Addis Ababa in order to attract park-and-ride users
for LRT transit system be a profitable system. The general study area LRT stations
accessibility is summarized in the following table5.2.
Table 5.2: Summary of study area stations accessibility
. Station name Road Section Number by Type Park-n- Kiss- Bus
Boulevard Collector Ring Road PAS SAS ride n-ride Stop
1 Lagahar 4 √ √ √
2 Stadium 2 1 √ √ √
3 meshualekiya √ √ √
4 Riche 2 2 √ √
5 TemenjaYazh 2 2 √ √
6 Lancha 2 2 √ √ √
7 Nifassilk 2 2 1 √ √
8 Nifassilk 1 2
9 Adey Abeba 2 2 √ √ √
10 Saris 2 2
11 Kaliti 2 2 √ √

(viii.) Population density

As observed in Figure 5.9, most of the Phase I stations are adjacent to zones with high
population density; High population density is one of the factors that affect ridership. High
population density implies that there is high ridership. Figure 5.9 shows population densities
in the Kebeles of Addis Ababa as provided by the Ethiopian Central Statistical Agency in
2007.

Population density was calculated for projected population and 3.2% Addis Ababa population
growth rate was used for the projection. And then, population density is the ratio of projected
population over kebele area in square Km.

Finally, population density information with other kebele attributes were tabulated and
mapped by using GIS for analysis purpose. The following figure 5.10 shows the calculated
and mapped study area population density within 3000m catchment of buffer area from

51
existing LRT stations. This density is grouped into five classes for analysis purpose.
Maximum population density is 73,000/sq.km. Thus, most of the stations in this study area
are accessible to walk-in traffic.

Figure 5.8: Study area Population density by kebeles

52
(ix.) Land use

Addis Ababa land use was categorized into ten groups to evaluate phase I LRT stations
integration with land use. As shown in Figure 5.11 below, most of the stations, especially
downtown, are located at the existing high density mixed residential and commercial areas.
This will help the stations attract large pedestrian riders. Areas of medium and low
population density will be converted to high density mixed use resident area, according to
city master plan. Hence, the AALRT stations will be more accessible to pedestrian users.

Figure 5.9: Study area Land use and LRT stations

53
 CHAPTER 6

6 PHASE II STATION LOCATION SUITABILITY ANALYSIS

6.1 Development of Station Location Selection Criteria
This section discusses the determination of criteria and classification of factors for the
identification of LRT potential areas which were divided into four main criteria and eleven
sub-criteria: six sub-criteria under criteria one; three sub-criteria under criteria two and one
each sub-criteria under criteria three and criteria four. Table 6.1 shows the details. These four
criteria are: land-use compatibility; transportation connectivity/intermodal connection;
community characteristics and topography. The eleven sub-criteria are: industrial, residential,
commercial, educational, institutional and public gatherings all under the criteria of Land-Use
Compatibility; then, bus connectivity/proximity to bus station, proximity to parking buildings
and connectivity to road were under Transportation Connectivity or intermodal connection
criteria; population density was under Community Characteristics criteria; and finally slope
was under Topography criteria

There were four criteria and eleven sub-criteria in the form of eleven GIS-based layers
incorporated for site suitability evaluation for LRT station. These criteria and factors for
identifying the potential station sites were acquired from literature assessment, previous
works and from planning, development and design professionals.

Based on the literature review and best practices, a series of guiding principles were used to
determine suitable sites and these were:
• The site should be at least quarter mile away (spacing distance) from the nearest
station in either direction;
• The site should be within 30m of commuter rail tracks;

For station locations that must accommodate park and ride facilities, locate them within a
five-minute walk of the platform (200m) distance but not directly in front of the station;
locate utility structures so as not to preclude redevelopment of prime station proximate sites.

A criterion is a basis for a decision that can be measured and evaluated. It is the evidence
upon which a decision is based. Selecting or formulating criteria that need to be fulfilled in
order to make the right decision is one of the difficulties in multi-criteria evaluation. The
related factors were grouped based on the four criteria. In this process, the data of all the

54
selected factors were kept, displayed and managed. Table 6.1 depicts Criteria developed for
station location.
Table 6.1: Criteria Developed for Station Location

Major Criteria Sub-Criteria

• Industrial
• Residential
• Commercial
Land-Use Compatibility • Education
• Institutional
• Public gatherings
• Bus connectivity/ Proximity to bus station
Transportation connectivity or Intermodal
Connection • Proximity to parking buildings
• Connectivity to road

Community characteristics • Population Density

Topography • Slope

6.1.1 Classification of criteria maps and data analysis

With regards to the acquired information, there were eleven important factors in the form of
six GIS-based layers incorporated for suitability analysis for LRT station sites. These factor
maps were overlaid together for final suitability evaluation of station sites. During the process
however, the data of all the selected factors shown in Table 6.1 were retained, displayed, and
managed individually. Because the factors have different scales of measurement, they cannot
be compared with the raw scores determined by experts.

Therefore, in order to allow comparability, the factor maps were standardized.

Standardization allows comparison of criterion scores within one alternative. In order to
standardize, the raster features of all the factors were reclassified into a common scale range.

6.1.1.1 Land-Use Compatibility

Integrate transportation and land use to optimize transportation investment and create an
accessible, efficient, and urban form. This criterion qualitatively evaluates the
appropriateness of the station in relation to its surroundings for compatibility with area land
use plans and general public acceptance.

55
It accounts for noteworthy construction, noise and light pollution impacts. It also considers
the potential for future transit-oriented development near the station and the impact a station
might have on area parcels.

The Addis Ababa Land Use Map of 2015 was classified into different classes of land cover.
For pair-wise comparison purpose, it is grouped into six classes: Residential, Educational,
Institutional, Commercial, Industrial and Public Gathering places. Figure 6.1 shows detailed
land use classification. Among the land use groups, residential is highly suitable for station
site because of attracting many public passengers. In this respect, high density mixed use
residential areas are highly suitable for LRT station sites. Figure 6.1 shows the Addis Ababa
Land use Cover as provided by the AASOID, 2015.

Figure 6.1: Addis Ababa Land Cover

56
6.1.1.2 Distance from roads

This criterion was classified based on Euclidean distance from collector road types. Areas in
the inner road buffer are best suitable for collector roads while ring roads (BRT lines) should
not be close to LRT line as it may create redundancy. Therefore, the outer areas outside of
buffers around collector roads are ranked as none potential. The result of the reclassified
Euclidean distance from the roads map is shown in figure 6.2.

Figure 6.2: Road Euclidean distance

6.1.1.3 Population density

Select rapid transit stations which have potential for densities sufficient to support rapid
transit service. Population data for the year 2007 were collected from the Central Statistical
Agency of Ethiopia. Population density for forecasted population for year 2025 was
calculated by using 3.2% growth rate and mapped. Areas with high population density are
highly suitable for locating stations. Figure 6.3 shows population density by kebeles of Addis
Ababa.

57
 Figure 6.3: Study area population density

6.1.1.4 Topography

Topography describes the surface shape and relief of the land. It refers to various landforms
(physical features) which represent the external shape of the earth (Tewodros, 2010).
Furthermore, elevation and slope should be considered when selecting site for stations. But in
this study, light rail lines run within the roads medians. Therefore, elevation factor was not
considered. Under these criteria, slope is one of critical factors/sub-criteria that can affect
station locations. The Addis Ababa slope map was classified into five classes in respect of
percentages where slopes 0 - 3% are highly suitable for stations. In the case of slope
calculation, the slope map of Addis Ababa area was developed by calculating the slope map
from the DEM (Digital Elevation Map) of the city. The slope calculator function is located in
the Arc toolbox of the ArcGIS software. The slope map was very essential in that it was used
to highlight suitable sites for the railway station. Figure 6.4 shows the slope map of Addis
Ababa.

58
 Figure 6.4: Slope Map of Addis Ababa

In this case, the reclassified slope map was acquired from the degree measurement unit for
station requirement. Flat landform is the most suitable for station. In view of the foregoing,
the 0-3 % slopes are ranked as highly potential; followed by 3-5% slopes as moderately
potential; 5-10% slopes as marginally potential; and above >10% with no potential. Desirable
maximum station-area grade or slope should be 0.5% and absolute maximum slope should be
2.5%. However, some existing conditions may require slopes exceeding the maximum 2.5%
in the case of on-street stations. The result of the reclassified slope map is shown in figure 6.4
above.

6.1.1.5 Proximity to other modes (access mode)

(i.) Proximity to Parking Building

Parking building is multi-storey car parking planned in the new Addis Ababa City Master
Plan and multi storey car parking is a building which is designed specifically to be for
automobile parking and where there are a number of floors or levels on which parking takes

59
place. Figure 6.5 shows proximity to parking building. For this study purpose, only parking
buildings that are found on 800m distance LRT lines are used. This is by considering
maximum distance that one can walk from parking building to nearest LRT station. The LRT
route from Summit to Laga-Tafo has no advantage of parking building since there is no
parking building within 800m catchment.

Figure 6.5: Proximity to Parking Building

(ii.) Proximity to Bus Stations/Stops

As seen in station accessibility analysis, the availability of other modes of transportation

around transit station is important for mobility around station area. The other modes of
transportation (Feeder Bus services etc.) facilitate easy movement of transit riders. Bus
service includes proximity of bus stations and stops to LRT station. Bus station in this case
includes intra-city bus station and intercity bus station. From figure 6.6, we can see that bus
stations are accessible to only some stations.

60
 Figure 6.6: Proximity of LRT stations to Stations

6.1.2 AHP Model Development

As stated by Cheung et al., (2002), AHP goes through the following three stages and figure
6.6 shows the four levels AHP model as developed in this research.
(i) Constructing the hierarchy for criteria and sub-criteria;
(ii) Comparing the weight of criteria and sub-criteria; and
(iii) Evaluating the alternatives by computation of the relative weight
of criteria and sub-criteria.

61
 Figure 6.7: Four level AHP Model

In this research, the fourth level is not included since, GIS is used in selecting suitable station
sites. If there were candidate stations used, AHP would have prioritized these station
alternatives.

6.1.3 Criteria Weights and Class Weights

This is determined based principally on the expert's opinions.

When the AHP method is applied to solve spatial decision problems in a GIS environment, it
is called spatial AHP method (Siddiqui et al., 1996). In the GIS database, the attribute factors
are represented by map layers and contain attribute values for each pixel in a raster data
format (Kiker et al., 2005).

AHP provides a structural basis for quantifying the comparison of decision elements and
criteria in a pair wise technique (Laskar, 2003). Once the pair wise matrix is made, relative
weights are calculated using pair-wise comparisons.

6.1.4 Pair-wise Comparisons

Pair-wise comparison of the criteria based on nine points Saaty AHP scale which is shown in
table 6.3 below was done by the experts and it was entered into 6 X 6 matrix as shown in
table 6.2. Their judgments on sub-criteria are shown in the tables 6.3 – table 6.9.

62
 Table 6.2: Relative Importance of each criterion

Table 6.3: Saaty nine point scale

6.1.5 Factors pair-wise comparison

Both criteria that are relative importance comparison and factor rating were done by experts.
For example, criteria that are extremely suitable for locating light rail transit station relative
to other criteria, will be assigned a value of 5; criteria that are partially suitable will be
assigned a value of 3; and criteria that are the less suitable for station sites relative to the
other criteria, will be assigned a value of 1. Tables 6.4 – 6.10 shows details of pair-wise
comparison.
Table 6.4: Five-point scale for pair wise comparison of sub-criteria scoring:

Score Class
1 Less suitable
2 Suitable
3 Moderately suitable
4 Highly suitable
5 Extremely Suitable

63
 Table 6.5: Category, land use, score and class

Category Land use Score Class

1 Residential 3
2 Industrial 3
3 Commercial 5
4 Education 4
5 Institutional 3
6 Public gatherings 5

Table 6.6: Category, slope, score and class

Category,
Range
range, score and
class Category Score Class
1 0 - 4.0 5
2 4.0 - 6.0 2
3 6.0 - 8.0 1
4 8.0 - 87.0 1

Table 6.7: Pair-wise Comparison of Distance from road factor

Category Range Score Class

1 0 - 100m 5
2 100 - 200m 5
3 200 - 400m 3
4 400 - 600m 2
5 600 - 800m 1

Table 6.8: Pair-wise Comparison of Distance from LRT Line factor

Category Range Score Class

1 0 - 30m 5
2 30 - 50m 2
3 50 - 100m 1
4 100 - 200m 1
Table 6.9: Pair-wise comparison of proximity to Parking Building Factor

Category Range Score Class

1 0 - 100m 5
2 100 - 200m 4
3 200 - 400m 3
4 400 - 600m 1
5 600 - 800m 1

64
 Table 6.10: Pair-wise comparison of proximity to Bus Station Factor

Category Range Score Class

1 0 - 100m 5
2 100 - 200m 5
3 200 - 400m 3
4 400 - 600m 2
5 600 - 800m 1

6.1.6 Weighting Criteria

By using experts‟ judgment of criteria relative importance, the criteria weights were
calculated as shown in table 6.11 below and these criteria were used in GIS multi-criteria
spatial analysis process. The calculation was done by using excel AHP template.
Table 6.11: Calculate criteria weights

n=6
RI = 1.24
CI = 0.032
CR = CI/RI = 0.025 which is less than 0.1 and the above result is acceptable.

65
6.1.7 Consistency Check
AHP allows some small inconsistency in judgment because of human imperfection. It is
necessary to verify consistency after the gaining of weight values (Chen et al., 2010). The
inconsistency measure is useful for identifying possible errors in judgments as well as actual
inconsistencies in the judgments themselves. As Rozann W. Saaty (2003) stated,
inconsistency measures the logical inconsistency of one‟s judgments. For example, if one
were to say that A is more important than B and B is more important than C and then say that
C is more important than A, then he/she is not consistent. It is accepted that the inconsistency
ratio should be less than 0.1. If inconsistency ratio is greater than 0.1, the judgment should be
revised. The ratio scales are derived from the principal Eigen vectors and the consistency
index is derived from the principal Eigen value. As provided by (Chen et al., 2010), the
consistency index (CI) and consistency ratio (CR) are depicted as equation (1) and equation
(2) below:

CI = (λmax-n)/ (n-1)........................... Equation 1

Where,

n = Number of the Criterion,

CI = Consistency Index
λmax = The biggest Eigen-value of the Comparison Matrix.

CR = CI/RI....................................... Equation 2
Where,
RI = a constant corresponding to the mean random consistency index
value based on n given in Table 6.11.
CR = Consistency Ratio

According to Saaty (1980), if the value of the consistency ratio (CR) is less than or equal to
0.1, the questionnaire is considered acceptable. If the CR is greater than 0.1, the questionnaire
is not acceptable. In this regard, the results of the analysis in this research showed CR =
0.025<0.1 which is acceptable so, the calculated criteria weights are used as GIS input data to
identify suitable station locations. Table 6.12 shows the consistency index value.

66
 Table 6.12: Consistency Index value

6.2 Results and Discussions

6.2.1 Multi-criteria preparation for GIS
The analysis that carried out in this project included proximity by buffering, overlay analysis.
The following results were obtained:
• A digital reclassified map showing spatial distribution of Phase II LRT Line;
• A Geodatabase containing various feature classes used in the study; and
• A suitability map.

Input data for GIS are: slope, population density, and distance from roads, land-use,
proximity to parking building and proximity to bus stations. These criteria were mapped and
overplayed in GIS by using the criteria weights calculated from AHP. Data for station
suitable sites analysis are prepared as shown in the issuing pages. The suitability analysis was
carried out with the aim of highlighting areas that are best suitable for phase II LRT stations.

6.2.2 Geodatabase
Before any analysis was carried out, data were converted, edited and organized into a single
personal geoatabase. A personal Geodatabase was developed and named
AALRT_System_Gotidatabase.mdb. It was occupied with feature datasets such as LRT
stations, LRT network, road network, population, and land-use and study area. The
Geodatabase provides an easy and efficient way of organizing and retrieving data. The
created Geodatabase is shown in figure 6.8 below as structure of the created Geodatabase.

67
 Figure 6.8: Structure of the created Geodatabase

6.2.3 Euclidean Distance Calculation

The Euclidean distance operation describes each cells relationship to a set of sources. The
function gives the distance from each cell center to the closest set in the raster format data set.
This function was used to calculate distances that are distance from bus station and parking
building to the stations. The Euclidean distance output is a raster file format that has different
layers or concentric rings from the point of reference. The distance is given in linear units
meters, kilometers, miles etc. In this study case, Euclidean distance calculation was done for
distance from road and from LRT lines.

6.2.4 Buffering
A buffer is an area formed from a locus of constant radius from a given point or along a
linear feature. This analysis can be done on either points or lines and the result is always a
new polygon layer. These buffer polygons define the inside and outside of a region which are
typically distinguished by different codes in the attribute table. In this project, buffering was
done on all the bus stops/stations and also on all Multi-storey parking buildings aimed at
identifying proximate bus stops and parking building to LRT stations.

6.2.5 Rasterisation
Rasterisation (or vectorization) is the task of taking an image described in vector graphics
format (shapes) and converting it into a raster image (pixels or dots) for output on a video
display or printer, or for storage in a bitmap file format. And Rasterisation refers to
conversion from vector to raster data.

68
6.2.6 Reclassification
This is the process of assigning weights to different classes of a raster data set. This process is
aimed at showing how the different concentric layers are placed and how suitable are they. In
this project the datasets reclassified were slope, Euclidean distance of railway road, Euclidean
of LRT line, land use in raster, and population in raster format too. Values were assigned to
each layer on a uniform scale of 1-5. This scale is thus referred to as the suitability scale.
Since all the data was harmonized by this process, the suitability process was possible to be
carried out.

6.3 Criteria Mapping and Spatial Analysis

6.3.1 Distance from Road
The Euclidean distances of road network within 1500m catchment from LRT routes were
calculated with a standard cell size of 30. This is because bus and taxi can access stations
from 1500m radius. It was then reclassified into five classes with 1 being the less suitable
and 5 assigned the extremely suitable.

69
 Figure 6.9: Reclassified Road Euclidean Distance

6.3.2 Slope
Slope was generated from the 30 meter DEM. It was then reclassified into 5 classes. Areas
with gently sloping terrain were the most suitable while rugged terrain was the least suitable
zone for the railway station. The slope was calculated in % of raise.

70
 Figure 6.10: Reclassified slope

6.3.3 Land-use
The different land use types were converted to raster and reclassified into 5 classes with 1
being the least suitable while 5 being the extremely suitable land use type for laying out a
new railway station. Figure 6.11 shows reclassified Land use types.

71
 Figure 6.11: Reclassified Land use

6.3.4 Population
The population of the area of the study was first converted to raster by the conversion tool,
and then reclassified. Five classes were assigned, with 1 being the least suitable while 5 being
the extremely suitable location for routing a new station site. Figure 6.13 shows reclassified
population density map while, figure 6.12 shows population density reclassification window
in GIS.

72
Figure 6.12: population density reclassification window

Figure 6.13: Reclassified population Density

73
6.3.5 Proximity to bus stops
Existing and planned bus stations were buffered with multi ring buffering tool and then
converted to raster map. This raster map was reclassified into five classes with class 1 is least
suitable location and class 5 is extremely suitable location for station site. Figure 6.14 shows
the reclassified bus station proximity.

Figure 6.14: Reclassified Bus Station Proximity

Only the route extensions from Torhayloch – Sebeta National railway station and Kaliti –
Akaki have bus station proximity within 800m catchment area. The other extensions do not
have bus station proximity.

74
6.3.6 Proximity to parking building
Existing and planned multi-storey parking were buffered with multi ring buffering tool and
then converted to raster map. This raster map was reclassified into five classes with class 1 is
least suitable location and class 5 is extremely suitable location for station site. Figure 6.15
shows the reclassified parking building proximity. The Summit – Laga Tafo and Kaliti –
Akaki routes has no proximity to parking building.

Figure 6.15: Reclassified Parking Building Proximity

6.3.7 Weighted Overlay

This was the last stage in the data analysis. This process was carried out by combining all the
reclassified raster data sets and defining their respective percentages of influence. This
process was best carried out by the model builder in ArcGIS. All the relevant raster datasets
were assigned different influential weights. The total percentage influence has to add up to
100%.

75
 Figure 6.16: Weighted Overlay Window

6.4 Suitability Map

Suitability analysis in a GIS context is a geographic or GIS-based process used to determine
the appropriateness of a given area for a particular use. The basic premise of GIS suitability
analysis is that each aspect of the landscape has intrinsic characteristics that are in some
degree either suitable or unsuitable for the activities being planned. Suitability is determined
through systematic, multi-factor analysis of the different aspect of the terrain. Model inputs
include a variety of physical, cultural, and economic factors. The results are often displayed
on a map that is used to highlight areas from high to low suitability. (Dramowicz, 2005).

A GIS suitability model typically answers the question, Where is the best location? It
involves finding the best location for a new station or pipeline, a new housing development,
or a retail store. For instance, a commercial developer building a new retail store may take
into consideration distance to major highways and any competitors' stores then combine the
results with land use, population density, and consumer spending data to decide on the best
location for that store.

76
6.5 Suitable Location Identification
This was done using model builder. Figure 6.17 shows the GIS processes (geoprocess) in
suitability model building and Figure 6.18 shows station location suitability analysis model
developed by using GIS.

Figure 6.17: Model Geoprocess

Figure 6.18: LRT station site suitability model

The output of this model was the map in next page which describes suitable intermodal and
accessible light rail transit station locations that is optimized using GIS integrated with
Analytical Hierarchy Process (AHP). Figure 6.19 – 6.20 shows station site suitability map.

77
 Figure 6.19: Station site suitability map for Torhaloch to Sebbeta National railway station

This station site suitability map is generated by overlying the six criteria mentioned at the
beginning of this chapter. From the map, the black blue color shows sited that are extremely
suitable to locate LRT station while, white color depicts highly suitable site and so on. So,
while we locate LRT station location it is recommended to locate it at black color areas if
possible. If not, the second suitable area for station location is the white area and the third
suitable area is the white blue area. We can locate stations based on minimum and maximum
station spacing at the same time by referring to this suitability map.

78
 Figure 6.20 : Station site suitability map for route from Arada to Addisu gebeya
This station site suitability map is also generated by overlying these six criteria.

79
 Figure 6.21: Station site suitability map for route from Summit to Laga Tafo
In this case, only four of the six criteria were used because there are no bus station and
parking building in LRT route proximity of 800m. From the map, black blue area is
extremely suitable. Since were only four criteria used, this suitability map may not be
intermodal as the other maps.

80
 Figure 6.22: Station location Suitability map for route from Kaliti to Akaki

The purpose of the above maps is, to give an idea of potential station site during station
location planning. After we get potential station sites that is determined by integration of
multi-criteria or objectives, best station location can be selected by considering minimum and
maximum station spacing and pedestrian walkable distance. As explained in above sections,
this map is a result of GIS suitability analysis that process multi-criteria that have direct
impacts in locating LRT stations. Proximity to road, bus station/stop, high density areas,
areas of high ridership origins and land uses were among the criteria that have considered in
this geo process. Therefore, one can refer or consult this suitability map for selection of
optimal LRT station location.

81
Figure 6.23: general station suitability map

82
 Figure 6.24: Potential Intermodal Station location

Based on this suitability map, minimum and maximum station spacing, and proximity to
other access modes such as bus station, we can select intermodal LRT station location. All
stations along LRT routes may not be needed to be intermodal station. This is because not all
stations are at a location of mixed access mode availability. Intermodal stations that can
accommodate all possible access or arrival modes are necessary to be designed for active
transport system creation in this city Addis Ababa. That is why the study tried to identify
suitable areas along LRT II routs.

83
 CHAPTER 7
7 CONCLUSIONS AND RECOMMENDATIONS
7.1 Conclusions
The purpose of this study was to identify suitable station sites along imminent Addis Ababa
Light Rail Transit (AALRT) routes. The study presents an integrated approach of GIS with
AHP combination to assess the station suitability by matching the characteristics of an area
with those attributes most appropriate for stations. These integrated approaches were able to
handle complex and many issues related to accessibility.

The main contribution of this study was the identification criteria and factors of LRT station
by applying the hierarchical structure of AHP in geospatial environment. It was started by the
calculation of weighting and rating from the AHP analysis where experts were asked to
determine the relative importance of each criterion and factor. The determination of criteria
and classification of factors for the identification of station potential areas which were
divided into three main categories: land-use compatibility, transportation connectivity and
topography aspects. There were seven criteria in the form of GIS-based layers incorporated
for location suitability of railway station.

The study has been successful in developing a methodology that identified LRT station sites
using GIS and AHP techniques. Those criteria and factors are very important components in
achieving station development as an integral part of the sustainable LRT system planning.
Beneficially, the final outcome of this study was the identification of the sites which are best
suited for intermodal LRT station in Addis Ababa along phase two routes. The study also
tried to take advantage of Addis Ababa planned and under construction multi story buildings
in identifying potential intermodal LRT stations.

Based on the outcome of the analysis, suitable intermodal station sites are mostly located at
areas where multi-storey parking buildings and bus transfer stations exist. By taking
advantage of surface and multi-storey parking, central stations can attract park-and-ride users
in addition to pedestrians. This will help the Addis Ababa LRT system to compensate its
expenses on high construction and operation-related costs.

In conclusion, a result of this study appears practically useful for the development of
effective LRT station sites and facilities planning. Additionally, final output of this study

84
could be used for generating candidate station location and prioritizing to select best station
locations.

Finally, the application of the outputs of this research can be useful for managers, operators
and planners working in transport-related fields.

7.2 Recommendations
The criteria and factors developed in this study can be used in planning and designing of
station potential sites. The developed methodology can be used for future Addis Ababa metro
intermodal stations site planning where intermodal station is demanded.

Since, light railway transit station location planning involves complex criteria in creating
active, accessible and sustainable urban rail transit, further study should be done with the
implementation of community/stakeholders participation in developing site specific station
location selection criteria.

With respect to the techniques implemented in this study, the integration of AHP in GIS
techniques has been proven beneficial for supporting decision-making. The methodology is
useful for identifying priority areas for LRT stations and this integration of AHP/GIS
methodology developed in this research can also be used for railway and highway alignment
selection. While, GIS is the best tool that should be used in handling various railway-related
data which will be easily accessed, edited and analyzed for railway station and alignment
design and for railway operation management.

85
 CHAPTER 8
8 FUTURE RESEARCH AREAS
This study should provide the motivation for the continuation of research and future
investigation on sustainable development of railway station and alignment planning and
design in Ethiopia.

Since, railway station and alignment design is a multi-criteria problem, integration of

Analytical Hierarchy Process (AHP) for best decision making and Geographic Information
System (GIS) for handling, analysis and calculation of spatial data related to the problem.

By using station site suitability map developed above, we can select many potential LRT
station locations which will be further screened and prioritized to get best station locations
along the route. Selection of optimum station locations from potential station locations shall
take into consideration factors such as drainage, slope, cost and engineering works and traffic
flow of the area.

86
 Reference

Akkelies van and Egbert. (2012). Degrees Of Sustainable Location Of Railway Stations:
Integrating Space Syntax and node place value model on railway stations in the province of
North Holland‟s strategic plan for 2010‐2040. Eighth International Space Syntax Symposium
(p. 25). Santiago de Chile: PAPER REF # 8005.
Alexis M. Fillone, Saksith Chalermpong, Sony Wibowo and Erik Vitug. Application of
Discrete Choice Modeling to Access Modes of The LRT Systems. CODATU XIII.
Amtrak. (2013). Amtrak Station Program and Planning Guide. Amtrak.

Anastasia Loukaitou-Sideris, Dana Cuff and Harrison Higgins . (2012). Up In The Air: Urban
Design For Lrt Stations In Highway Medians.
Bankable Feasibility Study for Addis Ababa Light Rail Transit 1st Draft. Addis Ababa.

Byrne, S. A. (2008). Station Location and Parking Supply Planning for Passenger Rail Lines.
Denver.

China Railway Group Limited. (2009). E-W and N-S Line ProjectT Study Reports. Addis
Ababa: Addis Ababa Light Rail Transit.

Chong, C. T. (2011). Sustainable Railway Development Through Careful Planning, Design

and Implementation. Hong Kong.: JOURNEYS .

David Evans And Associates ,Inc. and Sera Architects, Inc . (2009). Portland to Milwaukie
LRT Station Area Best Practices Assessments and Recommendations. Portland: David Evans
And Associates ,Inc. and Sera Architects, Inc .
David Evans and Associates, Inc and SERA Architects. (2009). Portland To Milwaukie Lrt
Station Area Best Practices Assessments And Recommendations. Portland: David Evans And
Associates, Inc | Sera Architects, INC.

Development, U. D. (June 2011). Station Area Planning for High-Speed and Intercity
Passenger Rail. FRA.
FDOT. (2014). Intermodal Transit Station Study Report. Florida: Florida Department of
Transportation by Ghyabi & Associates.
Franco K T Cheung, udy Leung Fung Kuen and Martin Skitmore. (2002). Multi-criteria
Evaluation Model for Selection of Architectural Consultants. Queensland University of
Technology.

Gleave, S. D. (2005). Yorkshire & Humber Accessibility Criteria. Wakefield: Yorkshire &
Humber Assembly.

Group.Lt., C. R. (2009). Bankable Feasibility Study forAddis Ababa Light Rail Transit 1st
Draft. Addis Ababa: Ministry of Transport and Communications (MoTaC) Ethiopian
Railways Corporation (ERC).

87
GTZ. (2004). Land Use Planning and Urban Transport. Enviroment and Infrastructure Sector
project Transport Policy Advise.

Hakan GULER , Murat AKAD and Murat ERGUN. (2004). Railway Asset Management
System in Turkey: A GIS Application. Athens.

Lai, X. (2012). Optimization Of Station Locations And Track Alignments For Rail Transit
Lines. Xiaorong Lai.
Lyon Town Planning Agency and AACPPO Addis Ababa City Planning Project Office.
(2012). Coordination of the mass transit network and urban development in Addis Ababa.
Addis Ababa.

MA‟SOEM, C. P., & NG, C. P. (2005). The Development Of Model Estimation To

Determine Parking Needs At Lrt Stations In Suburban Area. Serdang Selangor: Proceedings
of the Eastern Asia Society for Transportation Studies, Vol. 5, pp. 877 - 890, 2005.

Mepham, D. D. Why Is Light Rail More Pedestrian Friendly Than Bus Rapid Transit? A
Review Of Australian Urban Transit Accessibility.
Metropolitan Council. (2012). Regional Transitway Guidelines. Twin Cities Region.

Northern New England Intercity Rail Initiatives. (2014). Station Site Assesment and
Guidelines. VERMONT.

Olowosegun Adebola and Okoko Enosko. (2012). Analysis of Bus-stops locations using
Geographic Information System in Ibadan North L.G.A Nigeria. Industrial Engineering
Letters ISSN 2224-6096 (print) ISSN 2225-0581 (online) Vol 2, No.3, 2012 .
(2011). Regional Transitway Guidelines Technical Report Station Spacing and Siting.

Regional Transportation District. (2009). Rtd Transit Access Guidelines. Regional

Transportation District.

Scott A. Burger and Bill Byme . Station Location and Parking Supply Planning for Passenger
Rail Lines. Denver.

Shahroodi, Keramatpanah, Amini shabnam, Shiri, Elnaz and Najibzadeh. (2012). Application
of Analytical Hierarchy Process (AHP) Technique To Evaluate and Selecting Suppliers in an
Effective Supply Chain.

Spencer Environmental Management Services Ltd. (2013). Site Location Study. Edmonton:
Spencer Environmental Management Services Ltd.
(2011). Station Area Planning . U.S. Department of Transportation Federal Railroad
Administration Office of Railroad Policy and Development.

Sutapa Samanta, Manoj K. Jha and Charles O. Oluokun. (2005). Travel Time Calculation
With GIS in Rail Station Location Optimization. Paper: # UC2181 , 2-8.
Tefera, T. (2013). Addis Overview. addis ababa.

88
Tesfaye, D. F. (2012). Integrating Public Transport Networks and Built Environment.
Stockholm.
Transport Planning Team. (2015). Transport and Road Network Planning. Addis Ababa.

Utah Transit Authorityt. (2010). Light Rail Design Criteria Chapter 3. Utah Transit
Authority.

89
 Appendix

Appendix 1: Questionnaire
Please take a few minutes to fill this criteria pair wise comparison on railway station site
suitability analysis. I am conducting this to feed my research work for the fulfillment of MSc.
degree in Railway Engineering. I am greatly thankful to you for your time and appreciate
your cooperation. Your personal information will be kept confidential and will be solely used
for my own study for the above mentioned research. Thank you for your participation.

Personal Information

Providing the following information is optional.

First Name: _______________________________________________________
Organization:______________________________________________________
Position/ Role: ________________________________City: ________________
Telephone: ____________________________ Gender: ______ Age: _________

Research title: LRT Station Site Suitability Analysis by GIS and AHP

In this research, the following criteria are considered in the selection of suitable Railway
station location. These factors are grouped in to two: Main criteria and sub-criteria.

The relative importance of the criteria and sub-criteria has to be determined by experts from
different professions related to the topic.

A 9 point intensity of relative importance scale is used for pair wise comparison of Main
criteria and a 5 point importance scale for sub-criteria pair-wise comparison.

Criteria
• Land use compatibility
• Residential
• Industrial
• Commercial
• Education
• Institutional

90
• Public gatherings
• Transportation connectivity/multimodal link
• Bus connectivity/ Proximity to bus, taxi stops/transfer station
• Proximity to parking buildings
• Connectivity to pedestrian walkways
• Connectivity to road
• Population density
• Slope
• Elevation

Fill table2 below. Use 1 – 9 scales to evaluate their relative importance.

• Compare and determine relative importance of the following main criteria by pair-wise
comparison and circle to the value you think in table2: below.

Use a 9 point scale relative importance comparison table 1: below and encircle the value 1 –
9 in table2:

Table1 1: A 9 point scale for pair wise comparison of Evaluation Criteria

Sub-Criteria comparison

• For this detailed pair wise comparison, use the following 5 point scale given in table3
below and assign a score to these sub-criteria by comparing pair-wisely according to the
5 point scale given in table3: below.

91
 Table2: a 5 point scale for pair wise comparison of sub-criteria scoring

Criteria that are highly suitable for siting light rail transit station relative to other criteria, will
be assigned a value of 5; criteria that are partially suitable will be assigned a value of 3; and
criteria that are lowest suitable for station sites relative to the other criteria, will be assigned a
value of 1.

Pair-wise comparison of sub-criterion to Land-use factor Category, land use, score and class
Category Land use Score Class
1 Residential
2 Industrial
3 Commercial
4 Education
5 Institutional
6 Public gatherings
Category, slope, score and class

Category,
Range
range, score and
class Category Score Class
1 0 - 4.0
2 4.0 - 6.0
3 6.0 - 8.0
4 8.0 - 87.0
Pair-wise Comparison of Distance from road Factor

Category Range Score Class

1 0 - 100m
2 100 - 200m
3 200 - 400m
4 400 - 600m
5 600 - 800m

92
Pair-wise comparison of proximity to Parking Building Factor

Pair-wise comparison of proximity to Bus Station Factor

Pair-wise comparison of proximity to Road Factor

93
Appendix 2: Full Model Running window

Appendix 3: GIS tools used to build model for suitability analysis

94
Appendix 4: Participated experts list

95
Appendix 5: Addis Ababa population

96
 DECLARATION
I hereby declare that the work which is being presented in this thesis entitled
“ACCESSIBILITY AND SUITABILITY ANALYSIS OF LIGHT RAIL STATION
LOCATION BY USING (AHP) AND GIS: CASE STUDY ON EXISTING AND
FUTURE EXPANSION OF ADDIS ABABA LRT RESPECTIVELY” is original work of
my own, has not been presented for a degree in any other university; and that all sources of
material used for the thesis have been duly acknowledged.

Gomeju Taye Abera ______________ _____________

(Candidate) Signature Date

Submitted to: Addis Ababa University, Addis Ababa Institute of Technology

97