CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Module 4
Land use
INTRODUCTION
Land use characteristics and transportation are mutually interrelated. The use of the term
land use is based on the fact that through development, urban space put up a variety of
human activities. Land is a convenient measure of space and land use provides a spatial
framework for urban development and activities. The location of activities and their need
for interaction creates the demand for transportation, while the provision of transport
facilities influences the location itself. Land uses, by virtue of their occupancy, are
supposed to generate interaction needs and these needs are directed to specific targets by
specific transportation facilities. The following diagram explains the transportation land
use interaction

Land use means spatial distribution or geographical pattern of the city, residential area,
industry, commercial areas and the space set for governmental, institution or recreational
purposes. Most human activities, economic, social or cultural involve a multitude of
functions, such as production, consumption and distribution. These functions are
occurring within an activity system where their locations and spatial accumulation form

134

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

the land uses. So, the behavioral patterns of individuals, institutions and firms will have
an impression on the land use.

Land use system

The essential components of the land use system in terms of land use transport modeling
are location and development. The urban land use is largely modeled by simulating the
mechanisms that effect the spatial allocation of urban activities in the city. A number of
other important economic concepts underpin land use transport models, serving as
proxies for the complex interactions and motivations driving urban location. Among
these are the ideas of bid rent, travel costs, inertia (stability of occupation of land),
topography, climate, planning, and size.

Transport system
The second major component of a land use transport model, simulated along side land
use is the transport system the traditional way of characterizing the transportation system
in urban simulation models is a four stage process. The process begins with modeling
travel demand and generating an estimate of the amount of trips expected in the urban
system .the second phase trip distribution allocates the trips generated in origin zones to
destinations in the urban area. The third phase is modal split. Here trips are apportioned
to various modes of transport. The four stage simulation processes concludes with trip
assignment module that takes estimated trips that have been generated, distributed and
sorted by mode and loads it on to various segments of the transport network.

Factors affecting transport land use relationship

1. Urban land development
2.

Dominance of private vehicle ownership

3. Context of land use and transportation decision making

4. Different time contexts for response.
CLASSIFICATION OF LAND USES
The representation of this impression requires a typology of land use, which can be
formal or functional as explained below:

135

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Formal land use representations are concerned by qualitative attributes of space such as
its form, pattern and geographical aspects and are descriptive in nature.
Functional land use representations are concerned by the level of spatial accumulation
of economic activities such as production, consumption, residence, and transport, and are
mainly a socioeconomic description of space.
Land use, both in formal and functional representations, implies a set of relationships
with other land uses e.g. commercial land use has relationships with its supplier and
customers. While relationships with suppliers will dominantly be related with movements
of freight, relationships with customers would also include movements of passengers.
Since each type of land use has its own specific mobility requirements, transportation is a
factor of activity location, which in turn is associated with specific land uses.
LAND USE AND TRANSPORTATION
The movement of people and goods in a city, referred as traffic flow, is the joint
consequence of land activity and the capability of the transportation system to handle this
traffic flow exactly like that of principle of demand and supply. There is a direct
interaction between the type and intensity of land use and transportation facilities
provided. Ensuring efficient balance between land use activity and transportation
capability is primary concern of urban planning. Land use is one of the prime
determinants of movement and activity i.e. trip generation which needs streets and
transport systems for movement. This will lead to increased accessibility which further
enhances value of land and land use.
Land

Supply

Transpor

Population

Potential
Demand
And
Trip
Distribution

Equilibrium

Modal Split
Demand

Location

136

Trip Assignment

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Different Land Use Models

The purpose of land use transport models is to assess the policy impacts in terms of the
implications of the future growth patterns on both land use and travel related issues .For
this purpose, several researchers have developed various models with different theoretical
backgrounds and data requirements. From the early developments of land use transport
models to the latest state of art, can be broadly classified into three categories
(i)Early models (ii) Intermediate era models (iii) Modern era models.

Early Land Use Transport Models

There are several techniques which are representatives of earliest efforts in the
development of urban development models and which continue to serve (either in
original or modified form) a great number of transportation studies .These techniques are
quite simple generally deal with aggregate relationships .These are developed primarily
for location of residential activities. In addition many of these techniques can be applied
without using computer or simple programs can be prepared for use on a computer .These
simple techniques are considered most practical use in smaller urban areas because they
require less time, cost and data.
1. The Activity Weighted Technique allocates activity growth in population to share of
the particular activity which already exists in the zone .This technique assumes that
the present trends continue and allocates activity growth in proportion to the present
share .Therefore, the zone with highest present share will be allocated with major
share in future. It is clear that existing size as a proxy for the future development
potential leads biased allocation. This technique is suitable for short term planning.
2. The Density Saturation Gradient Method (Hamberg, 1959) is based upon the
axiom that there are regularities in the in the activity distribution about the central
place. The Density Saturation Gradient Method (DSGM) can be used as a tool for the
analysis of existing land use structure and also for use in forecasting land use
structure. The forecast is basically a trend projection of the existing land use and
density structure in the region. The method is based essentially on the regularity of
137

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

the decline in density and the percent saturation with the distance from the Central
Business District (CBD).This method depends equally upon the relationship between
distance and present saturation. Though the DSGM is complete in itself, this
technique demands more subjective inputs and allows only for a cursory and limited
consideration of policy and other planning decisions.
3. The simple Accessibility Model (Hansen, 1959) is based upon concept that the more
accessible an area is to various activities and the more vacant land area has greater
growth potential. Thus growth in a particular area is hypothesized to be related to two
factors; the accessibility of the area to some regional; activity distribution, the amount
of land available in the area of development. This accessibility of an area is an index
representing the closeness of area to all other activity in the region .All the areas
compete for the aggregate growth and share in proportion to their comparative
accessibility positions weighted by their capacity to accommodate development as a
measure by vacant usable land.
4.

The Intervening Opportunities (Lathrop, et al, 1965) model, spatial distribution of

an activity is viewed as the successive evaluation of alternative opportunities for sites
which are rank ordered in time from an urban center .Opportunities are defined as the
product of available land and density of activity. This model presumes that the
settlement rate per unit of opportunity is highest at the point of maximum access. The
concept of an opportunity for a unit of activity involves both land and measure of the
intensity of use of that land.

5. The Delphi Technique is a methodology for eliciting and refining expert or informed
opinion .The general Delphi technique involves the repeated consulting with a group
of individuals as to their best judgment as to when or what type of an event is most
likely to occur and providing with them systematic reports as to the totality of
judgments submitted by the group. The responses of all participants are assembled,
summarized and returned to the group members, inviting them to reconsider. This
information and revised estimates may be circulated to the participants for additional

138

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

anlysis.The procedure varies considerably among specific applications but the
primary result is that it produces a consensus of the judgments of a majority of
informed individuals while avoiding the bias of leadership influences , face-to-face
confrontation, or group of dynamics. Group of participants, are expected to clarify
their own thinking and the final decisions, according to the theory, it will tend to
converge by narrowing the range of estimates in response to the most convincing
arguments. Delphi is likely to involve more time and expense than the conventional
methods of forecasting.
All the early models are often considered as low cost models using simple theories.
Early developments of land use theory are simple techniques without much
complexity. Each of them has a sound basis and provides a reasonable estimate of
land use. How ever they do not cater for interaction of many variables. Some of these
techniques have been improved later for much better modeling strategy. It may be
seen from the inherent theories of this group of models, there is a broad city-wide
philosophy which operates the model and then zonal allocations are derived by
proportioning. Each of these models appears logical for urban land use forecasting or
activity allocation.

Intermediate Era Models:

This was the golden era of developments in land use transport modeling. Although , a
special group of models like empiric model has been developed and applied, the most
wide group of models is lead by the work of I. S. Lowry(1964).There are many variants
of one or more of these models as applied to particular area..
1. The Empiric Model (Hill, 1965) developed for Boston Regional Planning Project is
designed to distribute or allocate exogenously supplied growth forecasts of activities
such as population and employment along the zones and sub divisions of the region
considered for the study. This process of allocation considers the local changes in the
quality of public services and transportation network as well as changes over time in
the local activities .Although this model deals with population and employment other
activities can be incorporated into the model.

139

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

2. The Lowry Model (Lowry, 1964) incorporated within its structure both generation
and allocation of activities .The activities which the model defines are population,
service employment and these activities correspond to residential, service and
industrial land uses. Some of the salient features of Lowry model are
a) It assumes an economic base mechanism where employment is divided into basic
and non -basic sectors. Basic employment is defined as that employment which is
associated with industries whose products are largely used outside the region,
where as the products of the service employment are consumed within the region.
b) It is assumed that the location of basic industry is independent of the location of
residential areas and service centers.
c) Population is allocated in proportion to the population potential of each zone and
service employment in proportion to market potential of each zone.
d) The model ensures that populations located in any zones dose not violate a
maximum density or holding capacity constraint is placed on each category of
service employment.
e) Lowry model relates population and employment at one particular time horizon.
3. Garin expressed the fundamental Lowry algorithm in matrix format (Garin, 1966)
.Using this notation, the iterative process used by Lowry to generation population,
serving employment was replaced by elementary matrix operation to obtain an exact
rather than an approximate solution. The Garin formulation does not comprehend the
constraints which Lowry imposed .Neither the maximum size constraint for
population serving employment nor the maximum density constraint for residential
development was included in the matrix operations.
4. Time Oriented Metropolitan Model (TOMM) was one of the first derivatives of
the Lowry model (Crecine, 1964).Some of the characteristics of this model are
a) The model was developed in an incremental form contrary to static equilibrium
form taken by Lowry.

140

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

b) It attempts to disaggregate the locating population into several populations into
several types .It was felt that by disaggregating the model, the explanatory power
of the model would be increased.
c) Limitation of the study area to within city boundary.
d) There are different versions of TOMM; the structure of the revised model is
basically the same as original model although the allocation mechanisms have
been made more realistic.
5. Wilson Model based on entropy maximization is a break through contribution in the
urban spatial allocation models. It has enlarged the frameworks of spatial interaction
models. Wilson offered solutions to several problem areas by using the concept of
entropy maximization to generalize the problems. The concept of entropy was
originally developed in statistical mechanics and later proposed as general,
information applicable to most systems. The derivatives and the introduction of
entropy to urban and regional theory can be found in (Wilson ,1970) and (Wilson,
1974).The focus of the model includes
a) different household income groups
b) different wage levels by location of employment
6. Projective Land Use Model (Goldner, 1968) as another family of Lowry derivative
models. Projective Land Use Model (PLUM) is designed to yield projections of the
zonal level distribution of the population, employment and land use within an area
based upon the distributions of these characteristics in base year, coupled with a
series of simple and intitutively appealing allocation algorithm .There are different
versions of PLUM .Allocation incorporates auto and transit mode separately and
disaggregated local serving categories are allocated by different processes. The
allocation algorithms are derived from original Lowry model. This model can distinct
both basic and local-serving employment. The allocation function used in the model
has two components,
a) The first component is the probability of making a trip for a given trip purpose
of particular length

141

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

b) The second component is the measure of attractiveness of the destination
The total PLUM model is divided into four phases: initial allocation, revised
allocations of incremental employment, reallocations and increments, projections
The outputs of PLUM consist of total housing units, residential population, total
number of employment residents, and total employment.
7. Hutchinsons Model (Hutchinson, 1975) is being presented as a asset of land use
transport equations, refining more on transport aspects .These equations are capable
of analyzing alternative development strategies in sufficient detail to allow their
transport and serving implications to be examined .A procedure for corridor traffic
assignment analysis has been described, which uses the transport demand estimates
produced by land use model.
8. Sarnas Model (Sarna, 1979) is essentially a land use transport model which was
developed for Delhi, which was first of its kind and nature for application in India.
The model is based on the iteratively solved version of the Lowry model which
consists of a residential activity allocation sub model and a population serving
employment serving sub model.
The model deals with

Disaggregation by socio-economic group

Disaggregation by spatial groups

Simplified calibration procedure

Modern Era Models:

1980s has seen a very interesting development in the area of land use transport modeling.
During the intermediate era, modeling of transport demand and supply has been enhanced
with a lot of innovative ideas. The land use / transport modeling also embraced them foe
better representation of demand and supply scenario in relation to location. Thus although
the basic allocation mechanism emanated from Lowry model was largely used in most
models., very complex developments on location process can be found in the models

142

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

proposed. A significant assimilation of all such developments was taken up by TRL(UK)
through a consolidation study reported in 1988.The ISGLUTI (International Study Group
On Land-Use/Transport Interaction) study refers to nine models developed originally for
different cities of varying sizes and they have been comparatively evaluated for all modal
features (Webster, et al, 1988). This has also been tested for geographical transferability.
Some of the new land use models like cellular automata are also discussed in the report
(Timmermans, 2003).
The relationship between land use and transport means that any policy, whether relating
specifically to land use development or to the provision of transport facilities, will
inevitably affect the other dimension though not necessarily on the same time scale.
1. AMERSFOOT was developed to represent Amersfoot, a Dutch town of population
about 180,000 populations with the intention of examining land use planning policies

It is a spatial interaction model of the entropy maximizing type originally

formulated by(Wilson, 1970),though the general structure is similar to the
Lowry model

It takes the distribution of employment as given and the number of newly

built houses of different types is exogenously determined for ach zone on
the basis of structure plans and building plans formulated by the various
municipalities.

The population is disaggregated into three income groups , and the model
recognizes four types of locational behaviour of household changing
a) Jobs but not home
b) Home but not job
c) Both home and job
d) Neither of them

This model allocates workers from the zones containing their work place to
residential zones which are chosen in accordance with zonal attractiveness and
an exponential function of distance between residence and work place

There is no modal split, transport network or calculation of generalized travel

costs, because the intention is to provide a simple model which makes only

143

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

light demands which makes only light demands on computer storage or time
and to concentrate on land use policies.
2. CALUTAS (Computer Aided Land-Use Transport Analysis System) has been
developed to forecast the future location of housing, industrial and commercial
activities and the land use and travel patterns, within a large metropolitan area .As
applied to Tokyo , it represent a huge population of some 28 million within an area
of 15000 sq. km. In the model land uses are classified into the following four types
according to their locational characteristics:

priority location type(e.g. large scale basic industries )

optional locational type(e.g. business areas , housing)

subsequent location type(e.g. neighborhood stores , schools)

passive location type(e.g. agricultural areas, forests)

The allocation and amount of use is determined a priori on the basis of an

existing development plan. Allocation of optional land uses is described
by three five models
a) The industrial location sub model
b) Business location sub model
c) Activity within each of the zones
d) Local land use sub model
e) The transport sub model

3. DORTMUND is part of a compressive model of regional development organized in

three spatial levels (Wegner, 1982).A macro analytic model of economic and
demographic development of 30 zones .A misanalysis model of intra regional;
location and migration decisions in 30 Zones. A micro analytic model of land use
development in any subset of 171 statistical tracts within Dortmund. For these many
number of zones, the model simulates the inter-regional location decisions of
industry, residential developers and households. The resulting migration and travel

144

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

patterns, the land use developments and the impacts of public policies in the field of
industrial development, housing and infrastructure. This is done by six models
a) The transport sub model (calculates work, shopping, service and education
trips for four socio economic groups)
b) The ageing sub model (computes all those changes of stock variables
which are assumed to result from biological, technological or long term
socio-economic trends originating outside the model)
c) The public programmes sub model (possess a large variety of public
programmes specified by the model user in the fields of employment,
housing, health, welfare, education, recreation and transportation)
d) The private construction sub model considers investment and location
decisions of private developers
e) The employment change sub modes (models intraregional labour mobility
as decisions of workers to change their job location in the regional labour
market
f) The migration sub models (simulates inter-regional migration decisions of
households as search processes on the regional housing market.
4. ITLUP (Integrated Transportation and Land Use Package) model contains both
location and transportation models and has been the subject of a long sequence of
development and application projects since 1971.The four principal models are
a. EMPAL (Employment Location). EMPAL forecasts employment location
in each five year simulation period as a function of access costs by
population in different income groups for each zone
b. DRAM (Simultaneous household location and trip distribution). The
household (residence) location model allocates households to the zones
using a modified version of the standard singly constrained spatial
interaction model. Allocation of households of different types then
depends upon this attractiveness and the access cost to employment to
different types .In current version of ITLUP new locaters , include a
separate model LANCON to calculate land consumption using a

145

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

simultaneous multiple regression formulation. Trip generation and
distribution are also calculated in DRAM simultaneously with household
location.
c. MSPLIT for modal split calculation. The trip matrices produced in DRAM
are split into trip matrices for each mode in MSPLIT using multinomial
logit formulation for the modal split calculation.
d. NETWK for trip assignment. The trips are then assigned to a capacity
constrained highway network in NETWK.
5. LILT (The Leeds Integrated Land-use/Transport Model) (Macket, 1983)
represents the relationship between transport supply (or cost) and the spatial
distribution of population, housing, employment, jobs, shopping and land utilization.
It is applied to a study area divided into zones, with an external zoning system to
ensure the closing of the spatial system (Macket 1974).The main use of this model is
to allocate exogenously specified totals of population, new housings and jobs to zones
taking into account the existing land use pattern and the cost on travel and any
constraints on land use.
6. OSAKA has been developed to investigate the evolution of land use patterns in an
area where the land market is complex so that it was considered essential to
incorporate mechanisms which simulate the market and studies the land values.
OSAKA is an example of linear regression model in the EMPIRIC tradition.
Primarily it was considered to study land use effects rather than transport, the impact
of transport changes on urban development is of interest. The model provides no
predictive representation of either modal split or travel patterns.
7. SALOC (Landqvist and Mattsson, 1983) (Single Activity Model Allocation) is an
important model related to Herbet-Stevens model in that it maximises an objective
function and in general total interaction cost plays a major role. However, there are
certain deviations .It does not assume that all households wish to strictly minimise

146

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

their transportation cost, it contains other components such as neighborhood density
and infrastructure cost in the objective function.
a) SALOC allocates total population rather than net population growth
and specifies the cost of expanding the infrastructure. However,
constraints can be applied to the future population residing in the
housing cost of the base year, so that, in effect only the net population
growth allocated.
b) One of the main aim of SALOC is to identify short term developments
which will keep a maximum number of good options open for longer
term future when the prevailing condition may change.
c) It uses work trip pattern to accessibility of zones.
d) It does not calculate interzonal trip matrix, instead calculates a
composite travel time (cost) index for each zone as a weighed average
of travel times (cost) to the given set of work places and service
centers.
e) Basic philosophy is to provide a method of assessment which
integrates multiregional development with urban analysis, land use
with transport and normative planning with individual behaviour.
8. TOPAZ (Technique for Optimal Placement of Activities in Zone) is a general
planning model which allocates activities to zones of the study area in a way which
optimizes some weighted objective function of the cost establishment, on the other
hand the cost of operation which depends on the accessibility of the other.Special
applications to land-use/transport interaction studies, facility location and facility
layout. Paths between locations e.g. roads, rail network; costs (and benefits) of
location, eg. construction, operation, maintenance, Cost of interaction paths, e.g.
transportation costs, time periods for staging of developing or change. TOPAZ treats
the city as a system and the basic components of the city and their interactions are
matched by the components and the interactions of the model which assigns activities
to zones and the interactions to a network of flows.

147

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

9. The MEPLAN Model was developed by Echenique and Partners through a series of
studies in different countries in the world. It started with a model of stock and
activities followed by incorporation of a transport model developed for Santiago,
Chile, in the incorporation of an economic evaluation system for Sao Paulo, the
representation of market mechanisms in the land use model for Tehran the
incorporation of an input-output model again for Sao Paulo and the more compressive
model developed for Bilbao.
At the heart of the system is an input-output model to predict the change
in demand for space .A spatial system is used to allocate the demand to spatial zones,
using random utility concepts .An equilibrium model is derived by solving all the
equations, subjected to constraints .Given transport demand type and flow, the
transport model predicts modal split and assignment, with adjustment for times for
capacity constraints. Again random utility concepts are used in the transport model.
Information about costs , travel time due to congestion, etc are fed back into the land
use model to provide time lag measures of accessibility, (Hunt, 1994). Echenique
(Echenique, 1994) used the model to simulate the effects of urban policies.
10. The TRANUS integrated land use and transport-modeling system was developed to
simulate the probable effects of applying particular land use and transport policies
and projects and to evaluate their social, economic, financial, and environmental
impacts. A detailed explanation can be found in (De la Barra, 1989). Tranus has a
land use or activity model and a transport model. It is assumed that activities compete
for real estate, resulting in equilibrium prices, but also by accessibility, generated by
transport system. The location of activities is modeled in the land use system .The
transport model uses travel demand as input and assigns it. The land use model
generates a set of matrices of flows representing potential transport demand .The
purpose of the transport model is to transform potential demand into actual trips and
to assign these to the transport supply options.
11. MUSSA and RURBAN developed by (Martinez,1992) and (Martinez,1997) received
some interest because of spatial allocation of land uses is handled using a bid function

148

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

.The model is not a fully integrated model, but can accept as input the total demand
(growth) from the households and firms and a transport model. Central to the model
then is to predict the location of households and firms and the resulting rents.
Ellickson (Ellickson, 1981), showed that the spatial probability distribution obtained
from the bidding function is identical to the probability distribution obtained by the
maximization of individuals (consumers) surplus, emphasizing the equivalence of the
bid and choice approaches, given the traditional set of assumptions.
12. The model DELTA was developed by David Simmonds Consultancy, MVA
Consultancy and the Institute of Transport Studies, Leeds during the period 19951996. Consequently, it is not an integrated package, but a link of separate models.
Input to the land use model is that accessibility from each zone to alternative
destinations for each variety of purposes. The model predicts the location of activities
that are mobile as a function of accessibility, transport-related change in the local
environment, area quality and rent of space.
13. The initial design of the UrbanSim model was founded by the Oahu Metropolitan
Land Use Model as a part of larger effort to undertake the development of new
travel models. The project involved the development of a travel model system based
on modeling tours rather than trips. This model was further elaborated in 1996 when
Oregon Department of Transportation, launched the Transportation and Land Use
Model Integration Project (TULMIP) to develop analytical tools to support land use
and transportation planning. The model was extended and the prototype was
implemented. The model was calibrated for a case study in Eugene-Springfield. Later
the dynamic aspects of the model were calibrated and the model was applied in Utah
and Washington (Alberti and Waddell, 2000; Waddell, 2002).
14. The Integrated Model of Residential and Employment Location (IMREL) were
developed in connection with office of Regional Planning and Urban Transportation
of Stockholm (Anderstig and Mattson, 1998; Boyce and Mattsson, 1999). The model
starts with the total number of households and the total number of workplaces given

149

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

at the regional level. These are not predicted as a part of the model but exogenously
given these totals are then distributed across a system of zones through a process of
interactions between the residential and an employment location sub model. These
sub-models use as input data, among other things, travel times and travel costs
between zones by available models of transport as calculated by a traffic assignment
module of a linked travel demand model.
15. Using similar utility concepts, the same group also developed the TILT model
(Eliasson and Mattsson, 2001).Unlike IMREL; this model is descriptive by nature. It
models how the households, workplaces, shops, service establishments would locate
and interact, without any claim that the aggregate behaviour is optimal.
16. Uplan (Johnston, et al, 2003) allocates the increment of additional land in user
specified discrete categories consumed in future years. The model allocates future
development starting with the highest valued cells .As the higher valued cells are
consumed, the model looks for lower-valued cells until all hectares of projected land
consumption are allocated. In a recent test application for the Sacramento region, plan
was linked to a travel demand model to include the effects of changing accessibility
measured in terms of logsum (user benefit).
17. Integrated Land Use Transportation and Environment (ILUTE)modeling system
which is under development by a consortium of researchers in Canada from the
universities of Toronto,Calgary,Laval and McMaster (Miller and Savini, 1998).It is
an activity based integrated land use and transport model which represents an
experiment in the development of a fully microsimulation modeling framework for
the comprehensive , integrated modeling of urban transportation-land use interactions
and among other outputs the environmental impacts of these interactions.
18. The model Ramblas is developed to estimate the intended and unintended
consequences of planning decisions related to land use, building programs and road
constructions for households and firms (Veldhuisen, et al, 2000). The model allows

150

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

the planners to assess the likely effects of their land use and transportation plans on
activity patterns and traffic flows. It can simulate population of 16 million people.
19. The Irvine Simulation Models, of activity patterns that closely resembles to the core
of the Ramblas model. One important difference however is that the model is based
on a classification of representative activity-travel patterns. Some key aspects of such
patterns are extracted from the data and used to simulate activity travel patterns in a
particular environment. More recently the group is exploring the use of multi agent
systems (Rhindt, et al, 2003).
20. ILLUMAS is an integrated land-use modeling and transportation system simulation
project aims at a microscopic dynamic simulation of urban traffic flows into a
comprehensive model system, which incorporates both changes in land use and the
resulting changes in transport demand (Moeckel, et al, 2002).
21. Cellular Automata and Multi-Agent Models, in most of the cellular automate models
the transport component is weak. Typically a network is assumed but traffic flows are
not simulated. More recently, some scholar announced plans link their cellular
automata model with transport model. Central to these models is the use of cells that
can occupy particular states. Cell states may evolve according to transition rules,
which can either be deterministic or stochastic. Traditionally, dynamic process over
space were simulated for eight neighboring cells , but more recently applications
which use circular neighborhoods of a wider radius have been suggested (Engelen, et
al ,1997). In applications to land use patterns interaction mechanisms are usually
depicted in terms of distance decay functions. (Arentze, et al, 2003) have developed a
prototype of a system called Absolute.

151

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Land use /transport models

Predictive models

Optimizing models

Static models

Quasi-dynamic models

Entropy based
models

Spatial economic
models

Activity based
models

Classification of models

LOWRY LAND USE MODEL

The original Lowry was published in 1964 and since then several important extensions of
the original model have been applied to practical planning problems (Hutchinson, 1974).
The Lowry model conceives of the major spatial features of an urban area in terms of
three broad sectors of activity i.e. basic employment sector, the population serving
employment and the household sector. The basic employment is employment whose
products and services are utilized outside the study area.
With Lowry model, spatial distribution of basic employment is allocated exogenously to
the model while the other two activity sectors are calculated by the model by applying an
iterative procedure, until the constraints, which are maximum no. of household for each
zone and minimum population serving employment for any zone, are satisfied. The flow
diagram for this model is shown below.

152

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Exogenous Allocation of Basic Employment

Endogenous Allocation of Population

Endogenous Allocation of Population Serving

Employment

Check Constraints on Population and Serving

Employment

Sequence of Activities in the Lowry model

The model views the spatial properties in terms of:
1. Employment in basic industries
2. Employment in population serving industries
3. Household or population sector
Basic Employment: - employment in those industries whose products or services depend
on markets on external to the region under study.
The location of service employment is dependent on the population distribution of the
region.
Equation System
The above sequence of activities can be expressed in equation as follows.
P = eA -------------------------------------------------------------------------------(1)

e s = PB ------------------------------------------------------------------------------(2)

e = e b + e s ---------------------------------------------------------------------------(3)

153

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

where

P =row vector of population or household within each of the n zones

e =a row vector of the total employment in each zone

e s =a row vector of the population-serving employment in the zone

e b = a row vector of the basic employment in each zone

A = an nxn matrix of the workplace-to-household accessibility

B =an nxn matrix of the household-to-service center accessibility
The A accessibility matrix may be expanded as:

[ ][ ]

A = aij' a j --------------------------------------------------------------------------(4)

where

[a ]=an nxn square matrix of the probabilities of an employee working in i and living
'
ij

in j

[a ] =an nxn diagonal matrix of the inverses of the labour participation rates, expressed
j

either as population per employee, or households per employee

The B accessibility matrix may be expanded as:

[ ]

B = bij' [bi ]

where

[b ]= a nxn square matrix of the probabilities that the population in j will be serviced
'
ij

by population serving employment in i

[bi ] = a nxn

diagonal matrix of the population serving employment-to-population ratios.

The equations can be illustrated using the following example:

Total employment vector e = [126,177,64,216]
Basic employment vector e b = [100,150,40,200]
154

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

0.35
0.25
Journey to home function: aij' =
0.15

0.10

[ ]

0.50
0.30
Journey to shop function: bij' =
0.15

0.20

[ ]

0.25
0.45
0.20
0.25

0.30
0.35
0.10
0.25

0.20
0.20
0.35
0.20

0.10
0.15
0.40
0.35

0.15
0.10
0.25

0.20

0.15
0.20
0.40

0.45

0
0
0
0.80
0
0.80
0
0

Labour participation rate: a j =

0
0
0.80
0

0
0
0.80
0

[ ]

0
0
0
0.20
0
0.20
0
0

Service employment ratio: [bi ] =

0
0
0.20
0

0
0
0.20
0
The A and B matrices can be computed as:

0.28
0.20
A=
0.12

0.08

0.24
0.28
0.08
0.20

0.16
0.16
0.28
0.16

0.12
0.16
0.32

0.36

0.10
0.06
B=
0.03

0.04

0.05
0.09
0.04
0.05

0.02
0.03
0.08
0.07

0.03
0.02
0.05

0.04

The household vector may be calculated as:

0.28
0.20
[126,177,64,216]
0.12

0.08

0.24
0.28
0.08
0.20

0.16
0.16
0.28
0.16

0.12
0.16
= [95,128,101,142]
0.32

0.36

155

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

The service employment vector may be calculated as:

0.10
0.06
[95,128,101,142]
0.03

0.04

0.05
0.09
0.04
0.05

0.02
0.03
0.08
0.07

0.03
0.02
= [26,27,24,16]
0.05

0.04

Original total employment vector e = [126,177,64,216]

e b + e s = [100,150,40,200] + [26,27,24,16] = [126,177,64,216] ok!

Lowry-Garin Model
Garin proposed a formulation of Lowrys model which prevents the need for the iterative
solution to the equations described above.
Garin has proposed a formulation of the Lowry model, which obviates the need for the
iterative solution of to the equations. The following equations can be written:

Pb = eb A
e s (1) = P b B = e b ( AB) ---------------------------------------------------------------------(1)

P s (1) = e s (1) A = e b ( AB) A

e s ( 2) = P s (1) B = e s (1) ( AB) = e b ( AB)( AB) = e b ( AB) 2

Successive iterations will yield:

e s ( x ) = e b ( AB) x

P s ( x ) = e b ( AB) x A

Total employment and total population vectors are given by:

156

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

e = e b + e s (1) + ... + e s ( x ) += e b I + AB + ( AB) 2 + ... + ( AB) x + ...

P = P b + P s (1) + ... + P s ( x ) + ... = e b I + AB + ( AB) 2 + ... + ( AB) x + ... A

Garin has shown that under certain conditions on the product matrix AB will converge to
the inverse of the matrix ( I AB) and the resulting equations will be:

e = e b ( I AB) 1 ---------------------------------------------------------------------------(2)

p = e b ( I AB) 1 A -----------------------------------------------------------------------(3)

where
I =identity matrix

Garin argues that if this were not the case then an infinite amount of population serving
employment would be generated by a finite number basic employment.
Garin-Lowry model may be illustrated by the extension of the simple example given
above.

0.052
0.0480
AB =
0.0380

0.0392

0.0480
0.0496
0.0404
0.0464

0.0340
0.0364
0.0496
0.0456

0.0260
0.0260
0.0320

0.0288

which leads to:

( I AB) 1

1.0607
0.0567
=
0.0464

0.0477

0.0569
1.0585
0.0491
0.0522

0.0416
0.0441
1.0575
0.0534

0.0313
0.0313
0.3740

1.0342

The total employment vector will be: e = [126,177,64,216]

The household vector can be obtained as: P = [102,128,101,142]

157

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Sarnas Land Use Model

A critical problem in most Indian cities is the inadequacy of the transport infrastructure
which is further aggravated by the increasing demands for intra city travel due to rapid
growth in both population and employment. These demand based on travel forecasts led
to recommendation for high capacity facility, requiring large capital as well as operating
expenditure. The large investment required by the transport systems recommended for
Indian metropolitan cities have simulated an approach to land use transport planning
which attempts to minimize travel demands through the manipulation of land use (Sarna ,
1978).Dr Sarnas model considers Delhi as three districts as inner district, middle
district, outer district and different socioeconomic groups according to their income.

The Land Use Transport Model

The model, which is used by Dr Sarna, is a relatively solved version of Lowry activity
model, which consists of residential activity allocation sub-model and a population
serving sub-model.(Sarna, 1979).
The works to home linkages of the residential sub-model are calculated by the following
equation.

l ijkm = eik a k pr wkm h kj exp ik d im / h kj exp ik d ijm

j

. (3.1)

lijkm = The number of household (or persons) who are supported by employees of

income group k work in zone i and live in zone j and travel there by mode m.
eik = The total number of employees of income group k who works in zone i.

ak=The inverse of the activity rate of for income group k in terms of households
(or population) per employee.
pr wkm = The probability that employees in income group k will choose mode m

for the journey to work.

hj

= the

attractivity of zone j as a location for income group k households

158

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

i = the work zone specific parameter which reflects the influence that travel time
dij has on residential location selection by income group k employees
The number of household allocated to each zone are calculated from
p j = l ij
k

Where

km

(3.2)

pj

j=the

number of households (or persons) of income group k

allocated to zone j.
The home to service opportunities linkages of the population serving employment
sub model are calculated from

lijrkm = p kj b kr pr rkm sir exp jkr d im / sir exp jkr d ijm

i

lij

rkm

..... (3.3)

= the number of population serving employees of type r in zone j where the

service trips by residents of zone j are performed by mode m.

pj

= the number of households in income group k allocated to zone j by the

residential sub model.

r

s i = The attractivity of zone I for the location of type r service employment used in the

previous iteration of the service employment sub-model.

The home based work and service trip tables associated with the activity
allocations calculated by the above equations may be calculated by multiplying
equation (3.1) and (3.1) by the appropriate trip generation rates. Equation (3.1) and
(3.3) rely on trip end type modal split estimation in that the socio-economic
characteristic of trip makers are assumed to dominate modal choice decisions. Modal
split Probabilities that are specific to each i-j pair for each socio-economic group may
be substituted readily into the above equations.

159

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

START

SELECT TYPE
OF DETERANCE
FUNCTION
FIX jk

FIX BEST ik AND jk

SELECT BEST jk

RUN MODEL COMPARE

SIMULATED AND
OBSERVED TRIP
LENGTH DISTRIBUTIONS
POPULATION AND
EMPLOYMENT
DISTRIBUTION BY
INCOME GROUP

SELECT ik

RUN MODEL WITHOUT

CONSTRAINTS ON
ZONAL HOLDING
CAPACITIES

COMPARE SIMULATED AND

OBSERVED WORK
TRIPLENGTH
DISTRIBUTIONS AND
POPULATION DISTRIBUTION
BY INCOME GROUP

TEST
GOODNESS
OF FIT

TEST
GOODNESS OF
FIT

COMPARE SIMULATED
AND OBSERVED
SERVICE TRIP LENGTH
DISTRIBUTIONS
AND EMPLOYMENT
DISTRIBUTION BY
INCOME GROUP

RUN MODEL WITH

NO CONSTRAINTS
ON ZONAL HOLDING
CAPACITY

?
SELECT jk

SELECT BEST ik

FIX BEST ik

160

TEST
GOODNESS
OF FIT

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Model Calibration
The general procedure used to estimate the () and () magnitudes for the model is shown
in the figure.3.1. The magnitudes of ( ) and () were varied until the model activity
allocations and simulated trip length frequency distribution were in general agreement
with the characteristics observed in the base year.
The goodness of fit of the model was assessed principally by a subjective appraisal of the
model residuals. While more formal calibration techniques have been proposed and used,
it was felt that these more sophisticated measures of the goodness of the fit of the model
could not be justified in this investigation .The following criteria were used in estimating
() and () magnitudes

A minimum total absolute error between the given and model simulated
household and employment distributions and the absence of any
systematic spatial bias in the model residuals.

Good agreement between the observed and simulated work and service
trip length frequency distributions in terms of mean trip lengths.

Behaviour of the Model

The value of the parameters obtained from this model is being represented in relation to
the socio-economic group and the spatial distribution of the study area. The following
table provides a comparison of certain characteristics of the various versions of the model
examined for the base year conditions when the residential sub-model operated in an unconstrained way. The information presented in the table 3.1 is the parameter values
calibrated independently for the two sub-models i.e. residential sub-model, population
serving employment sub-model, table 3.2 shows the Comparative Performance at various
levels of disaggregation. Flowchart shows the comparative performance of the model at
various levels of disaggregation without constraints .It includes the percent absolute
model error in allocating activities, the

161

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Parameter values by income group for each district (Sarna, 1979)

Income Group

District
Inner

Middle

Outer

Lower

Middle

High

0.040 ()

0.030

0.050

0.100 ( )

0.130

0.140

0.150

0.130

0.130

0.140

0.140

0.150

0.160

0.150

0.100

0.130

0.140

0.150

Comparative Performance at various levels of disaggregation (Sarna, 1979)

Level of Disaggregation
Spatial

Socio economic Disaggregation

DisaggModel

Aggregated

Outputs

regation

Income Group

Low
%Model

High

All

Low

18.7

13.3

34.1

35.1

32.7

27.6

22.2

0.973

0.974

0.920

0.866

0.847

0.953

0.945

12.9

17.3

28.6

29.1

27.8

28.4

22.5

0.984

0.973

0.916

0.915

0.917

0.916

0.953

Ratio 0.979

0.941

1.042

1.078

1.005

----

0.956

1.022

0.994

1.001

0.812

-----

0.987

Househol

Error

ds

Correlation
Coefficient
%Model

Employm Error
ent

Middle

Correlation
Coefficient

Work Trip
length

Observed/Simulated
Service

Trip

Length 1.014

162

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Ratio
Observed/simulated
Simple correlation coefficient between the observed and modeled allocated activity
vectors and the ratio of the observed to simulated work and the service trip lengths.

Strategic Land Use Transport Model for Madras Metropolitan Area

(MMA)
A Lowry type Land use model has been developed for the MMA region in order to test
alternative development strategies together with their transport implications for a horizon
year of 2011.This model of land use transport interaction is developed at the strategic
level, utilizing an aggregated system of 65 zones with compatible transportation network
for testing the development strategies (IIT Bombay, 1993).
Area of Study

The city has sprawled over 172 sq km. with a number of urban roads The urban
agglomerations there have been a lot of developments in the form of additions ribbon
developments along the principal transport corridors has extended further inspite of
efforts made to plan and guide the developments. This study is aimed at arriving at a
suitable land use transport development strategy for Model for Madras Metropolitan Area
as a whole.
Scope and Objectives

Madras Metropolitan Development Authority (MMDA) desired to have a very

comprehensive Traffic and Transportation Study (CTTS) to fit with in the new structure
plan under preparation. It has therefore has been to prepare the land use transportation
strategy for MMA as the first step before taking up a detailed CTTS for the year
2011.The goals are as follows

Development of transport network proposals to achieve increased and more equitable

accessibility to employment and education opportunities and induce optimum land
use.

163

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Increased efficiency in the use of resources and economy in the public funds.
Conservation of human and natural resources .In general, this should involve
minimizing the overall cost of transportation.

The scope of the study for achieving the above mentioned objectives will be that

The model disaggregated for service employment will simulate the population and
employment distributions for the study area within the alternative development
constraints set for future.

Transport linkages derived from population and service employment allocations

mechanisms will generate transport flow patterns for the network due to the
development policies

Alternative blends of transport and land use strategies will thus get evaluated on
the basis of likely and desirable trends of growth.

Land Use Transport Model

The model used for this study is based on Lowry model according to which two major
functions are given by
It relates three elements of the urban/regional system, population, employment and
transport and relate their interactions

It incorporates within its structure both allocation and forecasting procedures

It assumes an economic base mechanism where employment is divided into basic and
non-basic(service)sectors

The basic employment sector includes those economic activities, the produce of
which is utilized mostly outside the region e.g. manufacturing and other heavy
government offices, the state head quarters, national financial institutions, university
etc. All other are accounted as non-basic (sector population serving employment).

The model assumes that the basic sector, both its location and magnitude is controlled
exogenously.

The model then determines the level and location of population and service (nonbasic) within the region.
The notations used are
164

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

=Population multiplier (inverse of labour participation rate)

k =Service employment ratio by type = Ek/P
Ek =Service employment by type k
Eb = Basic employment
P = Population
Since the total employment is
E = Eb + Ek
The loop of generating service and population will produce the total employment and
population as follows
k

E = (1 k ) 1 , and
1

P = E b (1 k ) .
1

BASIC
EMPLOYMENT

OBTAIN SERVICE
EMPLOYMENT
ES=P

OBTAIN
POPULATION
P=Eb

Economic Base Mechanisms (IITBombay, 1993)

Allocation Mechanism

(A) Residential location is a function of employment location and the trip making
behaviour of the population. The basic employment is allocated to residential zones for
using a singly constrained gravity model.

165

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Tij = Ai Ei H j exp(cij )
Ai = [ H j exp(cij )]1

Ei=is the employment in zone i (initially it is the basic employment)

Tij=number of people working in zone i and located in zone j for housing
Hj=attraction variable
cij= travel cost between i to j to be obtained from the network (in this case travel time
between zones)
=deterrence parameter of the allocation function to be calibrated with respect to the
base year work trip matrix
Ai=is the balancing factor
(B) The model uses the second allocation mechanism to locate the service as a function
of the location population and travel time. The functional form is given by
S ij = B j Pj Fi exp( k cij )
B j = [ Fi exp( k cij )] 1

Where
Sij = is the flow of people from residential zone j to service zone i
Pj=is the population distributed to zone j by the residential allocation mechanism
Fi= attraction variable of service center at zone i.
Bj=Balancing factor
B

The total number of people demanding services in zone i (Si) is therefore as follows
S i = S ij
j

The level of service employment required for each zone is estimated using service ratios.
Thus the service employment located in zone i for different service categories will be
Ei = 1 S i
1

Ei = 2 S i
2

Calibration Mechanism
The model is to be calibrated on the basis of given land use and transportation data. Its
aim being to simulate the distributed population and employment in the study area

166

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

/region. Thus the three parameters , 1, 2 will be estimated to satisfy the observed land
use distributions and travel matrices.
While calibrating the model to base year observed data the model will try to match the
observed distribution of population and categorized service employments. Thus it will be
working with constraints to match the land use and for this reason; the constraint in
population location will be applied. Any violation in allocation of population by
exceeding observed population modifies the attraction variable so that the allocation in
the following iteration gets corrected as follows.
Hj*= Hj (Pcj/Pj)

Where Pcj=population holding capacity (for calibration this will be observed population
P

in base year.) obtained from residential land available and policy on development with
respect to density
The violations in service employment are considered at lower end in terms of viability
(minimum size) constraint. Service employments allocated to zones are checked for
minimum size .For those zones where it is les than allowable minimum , these are
provided zero allocations and total of their allocations relocated in remaining zones of
higher allocations
Eik* = 0
Eik* = Eik

for zones where Eik < Ek (min)

for zones where Eik > Ek(min)

Only after the land use constraint are fully met, the model enters the transport loop where
it tries to match the observed work trip and service trip distributions .If it fails to satisfy
the defined limit of error the deterrence parameter of each distribution work trip,
education trip, and all other trip) will be corrected /modified/improved and the model
proceeds for the next iteration. The model starts afresh from the land use allocation as the
deterrence parameters control the accessibility in allocation function. This procedure
continues till all constraint on location of population and employment as well as those
related to trip matrices is fully met.

Five stage Land Use Transport model

The five-stage land use transport model (Lyon, 1992) has its decisions taken on instead of
the conventional four-stage land use transport model, it is based on

167

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Destination

Transport mode

Route

Mobility and

Location

And all of these are considered to be interdependent.

Mobility: It is the defined as the number of trips made a person and also as the type of the
trips made by a person. This is estimated by linear trip generation models by using socio
economic variables and the accessibility (Keoing, 1975, Dalvi, 1976, Martin, 1976).
Location: The existing land use transportation models are very much complex (Webster,
1988).so the proposed approach uses Bid choice model which is to be fully and
consistently integrated with the transport models using an extended decision chain of 5
components.
Bid Choice Model
-land, markets, location, rent

5 stage Land Use

Transport Model for
Urban Planning and
Land Use

Transport Planning Models

- trip generation, trip distribution,
modal split, trip assignment.

Fig 3.5 Outline of the 5 Stage Land Use Transport Model(Martinez, 1992)
General assumptions:

The consumer takes decisions on location and travel to achieve maximum utility

Consumer is willing to pay to enjoy the benefits of higher accessibility

Accessibility measures are revealed by consumer preferences in transportation and in

economic framework.

Mode choice decision can be taken care of in an economic framework by re

interpreting user benefits, land rents, and long term advantages of the transport
schemes.
168

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Consumers are all possible buyers of urban land, including types of household and
firms and possible to take care of tastes and priorities of all members involved.

The land is sold in land lot units, the land lot units are described by their r cultural
environment and it is assumed that human beings cannot change the attributes such as
view, accessibility, etc at their will.

Population land
firms and land stock

Land use and rents

Spatial location

Accessibility
Mobility (Trip generation)
Balancing factors

Trip rate
Destination (trip distribution)
Mode choice
O/D mode
flows

Route costs and

flow
Route choice

Land Use Model Structure (Martinez, 1992)

Models capable of being implemented in developing countries

The ISGLUTI study has brought together most of the fully integrated land use transport
models in current use .The formulations of the individual ISGLUTI models are not
surprisingly highly dependent on the interests and backgrounds of their modelers, the
reasons for developing the models in the first place, the type of city which the model has
169

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

been applied, the type of data available and the policy questions to which they were
intended to provide solutions.
No single model can claim to embody all that is best in the current state of the art or to
represent a universally optimal arrangement of components or of the various levels of
aggregation of the main parameters, though naturally each model provides what the
modeler considered to be the best representation of reality within his own particular
constraints.
Nevertheless, most models offer considerable flexibility within their considerations and
could be modified fairly and readily if desired to suit to the other country conditions to

Cope with different types of data

To give more detail on different aspects

To deal with different policy questions

To provide different types of information for policy makers

Policy Areas addressed by the models

Although the individual models were developed with a specific purpose in mind, they
represent land-use and transport evolution in a very general way and their applicability to
particular policy aspects is usually much wider than the application originally envisaged.
The table 2.1 indicates the different policy areas which can be addressed, at some useful
level of detail, by the various models.

Key:
9 The policy is addressed by the model.
a The model represents distribution of the population
b The employment must be located exogenously
Models

AMMERSOFT
CALUTAS
DORTMUND
ITLUP
LILT
MEP
OSAKA
SALOC
TOPAZ

Policy areas
Housing Employment Retail Public
Land- Transport Taxation
Infrastructure use
9
9
9
a
9
9
a
a
a

b
9
9
9
9
9
9
9
9

c
9
9
9
9
9
9
170
9
9

9
9

9
9
9
9

9
9
9
9
9
9
9
9
9

d
9
9
9
9
9
d
d
9

e
9
9
e
9
9
9
E
E

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

c Not in ISGLUTI, but examined in another model

d Some transport policies can be addressed, but comprehensive information on trip
behaviour is not available
e Some taxation/financing schemes affecting transport costs and car ownership.

Policy testing
1. Population change and land use restrictions:

Policy tests concerned with population change (Webster, et al, 1988)

Test

Model results available

Population grows at 2% p.a

No restrictions on land use
With restrictions on peripheral land use

ACDLMOT
ACDLMT

Zero population growth

No restrictions on land use

ADLMT

With restrictions on peripheral land use

ADLM

2. Employment location policies

Policy tests concerned with employment location (Webster, et al, 1988)

Model results available
Test
Half of non service jobs moved from inner areas to outer
ACDLMOT
areas
Half of non service jobs moved from inner areas to
ACDLMOT
peripheral industrial estate
Non service jobs redistributed in proportion to
CDLM
population
3. Location of shopping facilities and financial inducements
Policy tests with shopping and financial inducements (Webster, et al, 1988)
Test
Location of new facilities
City centre shopping floor space halved
New shopping center equivalent to one quarter of city
centre floor space set up in accessible location
Financial inducements
Unlimited free parking for city centre shoppers
Free public transport to city centre shops

171

Model results available

CDLMOT
DL
L
L

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

4. Cost of travel

Policy tests concerned with cost of travel (Webster, et al, 1988)

Test
All travel costs up 50%
All travel costs up 100%
Car costs quadruple
CBD parking cost= travel cost
CBD parking cost=3 * travel cost
Public transport free
Public transport fares up 50%
Public transport fares up 100%

Model results available

ADLM
DLM
DLM
DLM
ADLM
DLM
DLM
DLM

5. Speed and network changes

Policy tests with speed and network changes (Webster, et al, 1988)
Test
Speed Changes
1.Speeds of all mechanized modes increased by 20%
2. Speeds of all mechanized modes reduced by 20%
3. Bus speeds increased by 20%, speeds of other modes
decreased by 20 %
4. Speeds down by 15% in inner areas, 25% in outer areas
Network changes
1.New outer orbital motorway, speed is 80km/h
2.New inner road ring, speed is 60km/h
3.New cross town transit line, speed is 40km/h
4. As per 3 with speed 60km/h
Car Ownership
1.Growth in car ownership no extra investment in
transport network
2. As per 1 but car ownership grows by 2% more slowly
3. As per 1 but car ownership grows by 2% more rapidly

Model results available

ACDLMOT
ACDLMOT
DLMOT
DLM
DLM
DLM
DLM
DLM
DLM
DLM
DLM

6. Economic climate

Policy tests concerning with Economic climate (Webster, et al, 1988)

Test
Employment cut by 20%, travel costs increased by 20%
All travel costs up by 50%
All travel costs up by 100%
All people placed in same group of disposable income

172

Model results available

DLMT
ADLM
DLM
ADM

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Key:
A-

AMMERSOFT

C-

CALUTAS

D-

DORTMUND

L-

LILT

M-

MEP

O-

OSAKA

T-

TOPAZ

Mechanisms to be considered

The criteria, which are to be considered in transferring the model from one place to
another depends upon many factors like

Employment Location

Residential Location

Car Availability

Competition for land

Time and Space

Representation of travel

Model construction

Reliability of the models

Reliability depends upon many factors like its transferability satisfying criteria, its
behaviour after implementation etc. Reliability does not necessarily increase with
complexity or disaggregation, though the models which are too simple and global cannot
hope to fully replicate a complex situation. If the various mechanisms are thoroughly
understood and the strengths of the are known then the added complexity resulting from
the inclusion of more detail is likely to be justified.

173

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

Urban Goods Movements

Introduction
The urban goods movement, consisting mainly of truck transportation has been given
very little attention in transportation-planning studies. But in recent years the transport
planners, freight carriers, shippers etc. has understood the immense need of including the
commodity movements in the urban planning.
Goods movement demands are created by the economic activities of production and
consumption. It involves serious thinking regarding identification of principle economic
units in an area and developing an understanding their internal structures. A lot of
understanding is required in this regard because vehicle demand analysis is much more
complex than travel demand analysis and involves factors like separate routes for goods
movements, location of freight terminals and segregation of goods into small assignments
for distribution within urban area. The simple conceptualization of economic activities
can be shown by the diagram below.

Freight Movement Problems

The following four major problems have been identified regarding urban freight
transportation.
I. The interaction between commodity flows and land uses.
II. The general efficiency and economy of goods movement.
III. The environment problems of noise and air pollution.
IV. The truck movement in CBD.

Factors Considered In Goods Movement Forecasting

The factors important in forecasting of urban movements of goods are following.
I. Changing patterns of urban developments and structures
174

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

II. Location of terminals and transfer points

III. Land use patterns
IV. Changing costs and economics of goods movements industry
V. Labour practices within the industry
VI. Technological innovations in goods movements
VII. Effects of govt. police, aids and regulations
VIII. Social and environment considerations
IX. Inter-industry transactions etc.

Classification Of Urban Goods Movements

The classification of urban goods movements can be done at three broad levels as
following.
I. The first level classification is based on the spatial pattern of demand. This may
be further divided into groups like goods movements between urban area and
external locations, Inter-industry goods movements within an urban area and
household-based goods movements within an urban area.
II. The second level classification is based on commodity type which can be
perishable and non-perishable commodity and other such classifications
depending upon the type of industrial product.
III. The third level classification is by consignment size which is usually expressed in
terms of weight of the consignment.
The following diagram on the next page shows the broad classification of urban goods
movements.

175

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

External Commodity Movements

The commodity movements to and from external locations are of two broad types i.e.
direct consignments which are mainly made by trucks and consignments via a freight
terminal which involves pick up and delivery components by trucks. The proportion of
these two types is affected by freight pricing rules. Other mode of travel may be airlines,
rail and ships etc. Depending upon the trip length, the particular mode is selected as
shown in the table below.
Trip length

Transportation type

300 miles

Road Transport

300 900

Rail Transport

900

Water Transport

Input-Output Table
It highlights the economic structure of the industry. It consists of direct requirement
matrix. Each column of this matrix shows the dollar value of the inputs that is required by

a particular industry, being shown at the top of that particular column, from other
industries in order to produce one dollar of total output.

176

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

The input-output can be extended to include other important sectors like warehousing,
retailing, commercial etc. The necessary technical coefficients required to connect these
additional sectors can be established with the help of survey.
The input-output table provides a broad view of the average economic characteristics of
various industrial sectors. The annual inputs in terms of commodity type to a particular
zone may be obtained from equation as
aje = [aef] pje

Where aje = a column vector of the cash value of annual consumption by

commodity type e by industry in zone j.
[aef] = The direct requirement matrix of the input output table for e input and f output
industries.
pje = a column vector of the cash values of the annual production of commodity e in
region j.

Input-Output Table
Output Sector
Construction

Input
Sector

Building

Other
than
building

Hardware,
construction
Materials

Fuels

Wholesale Trade
Shop
Machinery
and
equipment
Office and
fittings supplies

Wholesale
Trade
Hardware,
construction
Materials
Fuels
Shop and
Office
fittings
Machinery
equipment
and
supplies
Pulp, paper
and
its
products
Other
businesses

Direct Requirement Matrix (pp.412 Hutchinson )

177

Retail Trade
Pulp,
paper
and its
products

Other

M/
vehicles

Other
consumables

Deptt.
And
Variety
stores

CE -751, SLD, Class Notes, Fall 2006, IIT Bombay

INDUSTRY

INDUSTRY NO.

SECTOR

1- Print/Publishing

0.03

.0001

0.0003

0.0

0.0006

0.0005

2-Iron, steel Mills

0.0
0.0

0.0

0.0

0.0

0.01

0.14

0.127

3-Primary Metals
4- Structural Metals
5-Metal Stamping

0.0

6-Other Fabric Ind.

7- Wages and Salary

0.15

0.29

0.11

0.24

Leontief and Strout Model

They have proposed the following gravity type expression for estimating interregional
commodity flows.
e

tij =

pi a j

qij

Where
tije= the cash value of annual flow of commodity e from region i to region j
qije= empirically determined coefficient which characterized interregional flow of
commodity e.

Wilson Model
It is a modification of the Leontief and strout model. The model can be

expressed

by

the equation as given below.

pi a j exp( e cij )
e

tij =

e
j

exp( e cij )

Where e = a parameter that expresses the importance of transport costs Cij on the
distribution pattern of commodity type e.

178