Surveying Models in Urban Land Studies

Abstract

Modern urban regions are highly complex entities. Despite the difficulty of modeling every relevant aspect of an urban region, researchers have produced a rich variety of models dealing with the manifold processes of urban change. This article reviews the models discussed in the literature in order to understand the most important aspects in modeling for urban studies currently. It starts by making a comprehensive review of the sixty-four models mentioned in the literature, and then "zooms in," detailing only the dynamic models (considered at the forefront of the fifth generation of systems theory). The issues explored include each model's subject areas and main goals, scales of analysis, and technologies used. Building on the previous criteria, the article aims to present a guide to existent models and point out future directions toward new technology developments in urban studies.

Full text

http://jpl.sagepub.com/
Journal of Planning Literature
http://jpl.sagepub.com/content/27/2/139
The online version of this article can be found at:
DOI: 10.1177/0885412211430477
2012 27: 139 originally published online 24 February 2012Journal of Planning LiteratureElisabete Silva and Ning Wu
Surveying Models in Urban Land Studies
Published by:
http://www.sagepublications.com
can be found at:Journal of Planning LiteratureAdditional services and information for
http://jpl.sagepub.com/cgi/alertsEmail Alerts:
http://jpl.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://jpl.sagepub.com/content/27/2/139.refs.htmlCitations:
What is This?
- Feb 24, 2012OnlineFirst Version of Record
- May 29, 2012Version of Record >>
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

Articles
Surveying Models in Urban Land Studies
Elisabete Silva
1
and Ning Wu
2
Abstract
Modern urban regions are highly complex entities. Despite the difficulty of modeling every relevant aspect of an urban region,
researchers have produced a rich variety of models dealing with the manifold processes of urban change. This article reviews the
models discussed in the literature in order to understand the most important aspects in modeling for urban studies currently. It
starts by making a comprehensive review of the sixty-four models mentioned in the literature, and then ‘‘zooms in,’’ detailing only
the dynamic models (considered at the forefront of the fifth generation of systems theory). The issues explored include each mod-
el’s subject areas and main goals, scales of analysis, and technologies used. Building on the previous criteria, the article aims to
present a guide to existent models and point out future directions toward new technology developments in urban studies.
Keywords
urban simulation, dynamic models, urban land-use change
Introduction
Recent decades have witnessed the evolution of urban land
models from traditional modeling such as linear and mathe-
matical models to complex dynamic models, triggered by the
application of computer modeling in urban studies from the
mid-1950s (Voorhees 1959; Batty 2004). The great strides in
computer techniques, particularly since the 1970s, in the areas
of complex analysis and artificial intelligence (AI; Los 1973;
Tobler 1979; Batty and Longley 1994; Silva and Clarke
2005; Silva 2008b) have contributed to the evolution of urban
models from more deterministic ways of representing reality to
complex adaptive forms that adjust to actual conditions and
include more stochastic and heterogeneous characteristics.
Although substantial changes have occurred in urban modeling
during recent decades, there is a need to increase the interaction
between scientific fields (i.e., social sciences and natural and
physical sciences) and this requires a seamless integration of
methodologies. This need is particularly important in modeling
the behavioral study of land-use change, which requires aspa-
tial characteristics of social economic and social psychology
together with spatial analysis. As a consequence, recent years
have seen a greater development of techniques and urban mod-
els that try to overcome the modeling divide of spatial versus
aspatial dynamics in urban land-use change.
The increasing understanding of the functionalities and
potential application of the models being studied changed the
role of computers allowing the support of planning practice and
the adoption of participatory spatial planning tools (Pettit et al.
2008). While many kinds of models have been used by a host of
diverse professionals, there still remain issues in bridging the
gap between computer applications in labs and the practice
of urban studies. For instance, one of the ‘‘bottlenecks’’
commonly acknowledged in the literature is the adoption of
models by planning practitioners; as Vonk, Geertman, and
Schot (2005) pointed out, much effort has been made to
develop software tools for planning while little has been done
to support and evaluate their application.
Increasingly, model development requires interdisciplinary
knowledge and dedicated consideration of many aspects of the
model construction procedure. In a previous study (Wu and
Silva 2010), we reviewed the AI approaches and algorithms
that have been embedded in systems or models for urban stud-
ies. This article follows up from this previous study and
reviews current commonly used dynamic urban land models
and modeling development procedures that support the practice
of urban planning. In addition, based on the review of current
urban models, and the literature for future modeling develop-
ment, this article concludes with a discussion of the future
development of urban dynamic models.
Our article is organized as follows: after this introductory
part, the section Current Pracitce and Frontiers reviews a wide
spectrum of models using a set of benchmarks, and the section
An Overview of Existent Urban Models and a ‘‘Zoom In’’ to
Dynamic Models discusses urban dynamic models in particu-
lar. A set of concluding remarks ends the article.
1
University of Cambridge, Cambridge, United Kingdom
2
Harvard Real Estate Academic Initiative, Harvard University
Corresponding Author:
Elisabete Silva, University of Cambridge, 19 Silver Street, Cambridge, CB3 9EP,
United Kingdom
Email: es424@cam.ac.uk
Journal of Planning Literature
27(2) 139-152
ªThe Author(s) 2012
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0885412211430477
http://jpl.sagepub.com
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

Current Practice and Frontiers
In order to overcome the information gap that many practi-
tioners face when selecting a dynamic model, this article com-
pares a variety of leading urban land models currently being
used or under development that could be employed by planners
and scholars. In order to achieve this goal, we first define the
following six benchmarks in order to analyze and classify
urban/land-use models: (1) different modeling approaches,
(2) levels of analysis, (3) spatial scales, (4) temporal scales,
(5) spatial and aspatial dimensions, and (6) different planning
tasks. By understanding these models’ characteristics, we hope
to direct them to particular urban problems.
Benchmarks of Classification and Model Generations
Over the past few decades, the development of urban models
has increased exponentially. The variety of models makes their
classifications quite diverse and difficult; as a result, research-
ers have proposed many kinds of benchmarks for classification.
In this article, we will survey and update past classifications
and extend those in order to include more models and new
modeling approaches and functionalities. As a result, in this
article, we propose a new classification in order to survey as
many models as possible in a more comprehensive way.
Modeling approaches have evolved from traditional app-
roaches such as linear and mathematical techniques to ones that
include more complex, dynamic, and intelligent elements.
Researchers have proposed a variety of modeling approaches;
for example, Mitasova and Mitas (1998) outlined urban model-
ing approaches including deterministic, stochastic and rule-
based and agent-based modeling. Batty (2007) analyzed and
classified models based on agent-based modeling, cellular
automata, and fractals.
In this article, we take the modeling approaches as the first
benchmark of our models’ classification. The urban system is
the synthesis of interactions between the micro and the macro
levels; therefore, we consider the level of (micro or macro)
analysis as the second benchmark for our classification of mod-
els. However, because all the urban dynamic models incorpo-
rate different spatial and temporal interactions, we therefore
take spatial and temporal scales as our third and fourth bench-
marks of classification. Urban systems include spatial and
aspatial dynamics; this important aspect constitutes our fifth
benchmark: spatial/aspatial. Considering that models are
designed for different planning tasks, this article takes the plan-
ning task as the sixth benchmark.
Due to the variability of the models, the approaches taken
and the secrecy involved in some of the modeling strategies
(which, in some cases, make them almost black boxes), the
overall classification we present could be more exhaustive in
the description of the different models; nevertheless, with the
given information provided by the authors, the only option was
to include solely the evidence from literature (i.e., articles and
reports published that describe the models). Due to the length
of the article, we could not present each model in detail;
therefore, we chose the option of selecting and detailing models
that are representative of each benchmark (providing a list and
a brief description of all surveyed models in the respective
synthesis table for each benchmark).
The modeling literature suggests several model classifica-
tion schemes. When assessing this great variety of models,
many researchers have outlined categories and classifications
of urban land models (Batty 1976; Issaev et al., 1982, Stahl
1986; Briassoulis 2000; Verburg, Schot, and Veldkamp 2004).
Verburg, Schot, and Veldkamp (2004), for example, reviewed
land-use change models from different perspectives (micro,
macro, and cross level) and the level of integration. After
reviewing the literature, and according to their different charac-
teristics, methodologies, application areas, and modeling
approaches, we observed that models can be classified into
many types such as long-term and short-term models, loose-
couple and tight-couple, linear and unlinear, static and
dynamic, mathematical and computer simulation models.
The notion of ‘‘generations’’ of modeling to describe the
trend of model development was introduced by Abbott (1989)
and Cunge (1989). Each generation of models in urban studies
introduces some novelties and addresses specific problems of
its time. The first generation of computational models was
developed during the Post-War period, and was popular in the
1950s with the development of large-scale urban models. This
generation of urban models was based on applying such tech-
niques by treating the urban system as a static entity whose land
uses and activities were to be simulated at a cross section in
time and whose dynamics were largely regarded as trending
toward self-equilibrium. The Lowry model is the best known
of these models and predicts the location of urban residential
and service activities for given locations of basic industry. The
early models were in the tradition of comparative static analy-
sis (Lowry 1964). These models faced multiple challenges of
implementation and were largely abandoned by the end of the
1960s. Some researchers, Lee (1973), for example, criticized
these models, mentioning that they had almost nothing to do
with spatial structure.
The second generation of models was a result of the collapse
of the first generation. It is characterized by sector-by-sector
approaches rather than highly complex and integrated models
(the transportation sector, where many of these second-
generation models were developed, is a good example—mainly
deterministic models). Technique for Optimal Placement of
Activities in Zones (Brotchie 1969) is an example. The model
is a nonlinear programming model that allocates activities to
zones on the basis of maximum utility or welfare for the com-
munity. In addition, it was developed in the period of reevalua-
tion of modeling techniques and later of resource constraints on
urban energy and growth (Brotchie 1978).
The third generation of modeling focused on the develop-
ment of solutions for specific domain problems, and, as a
result, could only be apprehended by the modeler and highly
specialized users well trained over a long period. A good exam-
ple of a model of the third generation is Brotchie’s model
(Brotchie 1978). This marks an important period during which
140 Journal of Planning Literature 27(2)
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

computational capabilities, availability of data and expertise in
the different scientific areas started to catch up with develop-
ments in other scientific areas. Typical examples are some
sophisticated two-dimensional or three-dimensional (3D)
finite-difference numerical models for land-use change and for
particular land-cover phenomena.
The fourth generation of modeling has become much more
useful to a much wider range of end users by furnishing menus
of parameters. Yet, they do not address the core problem of
knowledge elicitation and transfer. CUF (Landis 1994), for
example, was one of the first GIS-based urban growth models
used to simulate regional and subregional growth and relies on
vector (polygon-based) data for its analysis.
The fifth generation of modeling systems is acknowledged
to have the features of integrating ‘‘weak’’ AI (Bethell 2006)
technology and computational hydrodynamics into a single
system and, also, to provide assistance for nonexperienced
users (Chau 2006; Chau and Chen 2001; Abbott 1989). CUF-
2 (2001) is a representative example in this generation; it made
significant changes to the original CUF model resulting in a
cell-based model that uses regression analysis to determine
land conversion probabilities.
Models’ Classification and Analysis
Benchmark 1: Different modeling approaches. Cellular auto-
mata, agent-based models, and fractals (which can be modeled
in the scope of some GIS applications or work as stand-alone
models) have been indicated as the main approaches to under-
standing cities (Batty 2007), and therefore are some of the most
promising methodologies in this new generation of urban
models.
Batty’s statement is confirmed by the increasing number of
dynamic models that use CA or agents representing the
dynamics in urban systems. Important modeling approaches
also include expert systems, particularly rule-based systems
(incorporating explicit decision rules, which allow model users
to specify how the model will behave) and GIS-based models
(where the model relies mainly on GIS and the empirical
knowledge of the users). Traditional mathematical/statistical
modeling approaches continue to be used very often for spe-
cific applications. Table 1 presents a classification of models
into six groups according to the above modeling approaches.
Benchmark 2: Different levels of analysis. As Verburg, Schot,
and Veldkamp (2004) described, models based on the micro-
level perspective are all based on the simulation of the behavior
of individuals and the upscaling of this behavior, in order to
relate it to changes in the land-use pattern. Many models can
be categorized in this camp, for example, Wagner and Wegener
(2007) described the ILUMASS project as being of micro-
scopic approach in detail. The macro-level models simulate
phenomena based on macroeconomic theory or the system
approach. These kinds of models emphasize the macro pro-
cesses of urban land-use change (LUC) and hardly consider the
micro-level interactions. We consider the other urban models
that do not have specific emphases on the macro or micro level
as the cross level urban models. Based on this line of argument,
we can classify models into three groups, as shown in Table 2.
Benchmarks 3, 4, and 5: Different scales, timelines, and spatial/
aspatial dimensions. Similar to the micro and macro perspec-
tives, the spatial and temporal scales are another kind of base-
line for models’ classification, but they focus on the scales of
the simulated phenomena in terms of the size of the application
area and the prediction time period. Many researchers refer to
spatial scale as a regional or local application of urban models;
in most cases, they differ according to the understanding of the
authors (take the UrbanSim model as an example, classified
Table 1. Models with Different Modeling Approaches
Benchmark of Classification: Modeling Approach
Classifications of
Models Models References
Mathematical/
statistical
models
POLIS Prastacos (1985)
PLUM Prastacos (1985)
STIT Nuzzolo and Coppola (2005)
GIS-based models CLUE Verburg et al. (2001)
CURBA Landis (2001)
METROSCOPE Larson, Cser, and Conder
(2000)
PECAS Hunt and Abraham (2003)
UPLAN Walker et al. (2007)
Cellular
automata-based
models
CVCA Silva, Wileden, and Ahern
(2008)
DUEM Batty, Xie, and Sun (1999)
LOV White, Straatman, and
Engelen (2004)
LEAM Deal (2001)
LUSD He et al. (2004)
SLEUTH Silva and Clarke (2002)
UED He et al. (2008)
Agent-based models
(ABM)
FEARLUS Polhill, Parker, and Gotts
(2005)
PUMA Ettema et al. (2005)
LUCITA Lim et al. (2002)
LUCIM Hoffmann, Kelley, and Evans
(2002)
SYPRIA Manson (2005)
FEARLUS Polhill, Parker, and Gotts
(2005)
Rule-based models CUF Landis (2001)
CommunityViz Kwartler and Bernard (2001)
INDEX Allen (2001)
Place3S Snyder (2001)
SAM-IM MAG
UPLAN Walker et al. (2007)
What if? Klosterman (2001)
Integrated models CVCA Silva, Wileden, and Ahern
(2008)
DG-ABC Wu and Silva (Forthcoming)
ILUTE Miller (2001)
ITLUP Putman (1983)
SLUDGE Parker and Najlis (2003)
UrbanSim Waddell (2002)
Silva and Wu 141
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

into the large-scale model group by Pinto and Antunes 2007,
but it is capable of simulating at a local scale such as that of the
community).
In a first approximation to proposing a classification that
groups models accordingly to spatial and temporal scales, mod-
els are evaluated in a very rough way through larger, intermedi-
ate or smaller scales in terms of the temporal scale. Therefore,
this study groups models by the spatial scales from local to
metropolitan or regional (or national) scales in Table 3, and
time phases from zero year to ten years (short term) to ten years
to fifty years (medium term) or more than fifty years (long
term) in Table 4. Due to the limitation of the information about
the urban models, it is difficult to produce the concrete defini-
tions about the spatial scales; as a consequence, the classifica-
tion in Table 3 is mainly based on models developers’
descriptions of models’ spatial scales.
Finally, in considering the spatial or aspatial characteristics
of models, we can categorize urban land models into three
groups (as is shown in Table 5): spatial-oriented models (focus-
ing on geographical patterns of LUC processes), aspatial mod-
els (focusing on the interaction of commodity demand,
production, and trade) and integrated models (concerning both
geographic and socioeconomic aspects).
Benchmark 6: Different planning tasks. Another possible clas-
sification approach is to classify models according to their
designed planning tasks, for example, Klosterman and Pettit
(2005) categorized some planning support systems (PSS) into
four groups according to their tasks: land-use/land-cover
change, comprehensive projection, 3D visualization, and
impact assessment. PSS are a subset of geoinformation technol-
ogies dedicated to aid those involved in planning to explore,
represent, analyze, visualize, predict, prescribe, design, imple-
ment, monitor, and discuss issues associated with the need to
plan (Batty 1995). Although models and PSS are not 100 per-
cent interchangeable, to some extent the models follow similar
generic planning tasks with PSS. For example, PSS are pri-
marily model-driven decision support systems aiming at
rationalizing planning. It aims at providing necessary support
Table 2. Models with Different Levels of Analysis
Benchmark of Classification: Level of Analysis of Models
Classifications of
Models Models References
Micro level CUFM Landis (1994)
ILUMASS Wagner and Wegener (2007)
ILUTE Miller et al. (2004)
IRPUD Wegener (1998)
METROSIM Anas (1982)
PUMA Ettema et al. (2005)
SIMPOP Sanders et al. (1997)
TLUMIP Weidner et al. (2006)
Urbansim Waddell et al. (2003)
Macro level CLUE/CLUE-S Verburg et al. (2001)
GEOMOD2 Pontius, Cornell, and Hall
(2001)
IIASA Fischer and Sun., (2001)
LOV White, Straatman, and Engelen
(2004)
LTM Pijanowski et al. (2000)
Cross level (or
multilevel
models)
BabyLOV White, Straatman, and Engelen
(2004)
WiVsim Spahn and Lenz (2007)
Table 3. Models with Different Spatial Scales
Benchmark of Classification: Spatial Scales
Classifications of
Models Models References
Regional scale (large
scale)
GEOMOD2 Pontius, Cornell, and Hall,
(2001)
LEAM Deal (2001)
METROPILUS Putman and Shih-Liang
(2001)
SPARTACUS Lautso (2003)
SAM-IM MAG
TRANUS de la Barra (2001)
Metropolitan scale DRAM Putman (1995)
ILUMASS Wagner and Wegener
(2007)
PUMA Ettema et al. (2005)
Local scale CARLOS Bru¨cher et al. (2000)
IFDM Mensink and Cosemans
(2005)
CommunityViz Kwartler and Bernard
(2001)
Place3S Snyder (2001)
Multi scale CLUE Verburg et al. (2001)
CVCA Silva, Wileden, and Ahern
(2008)
DG-ABC Wu and Silva (Forthcoming)
Environment
Explorer
White and Engelen (2000)
SLEUTH Silva and Clarke (2002)
SOLUTIONS Echenique (2004), Yin
(2005)
UrbanSim Waddell et al. (2003)
Table 4. Models with Different Temporal Scales
Benchmark of Classification: Temporal Scales
Classifications of
Models Models References
Long term FEARLUS Polhill, Parker, and Gotts
(2005)
LUCIM Hoffmann, Kelley, and Evans
(2002)
MEPLAN Echenique and Hunt (1993)
UGM Clarke and Gaydos (1998)
Medium term Agent-LUC Rajan and Shibasaki (2001)
CLUE Verburg et al. (2001)
Short term SelfCormas
experiment
D’Aquino et al. (2003)
SAM-IM MAG
142 Journal of Planning Literature 27(2)
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

to systematically analyze the information, findings and formu-
lation of the problems, structure and formulate the alternative
options, policies, scenarios and plans, assess and evaluate their
impacts (consider the objectives of the relevant stakeholders)
and finally, select and recommend a proper solution (Power
2004). Based on Klosterman and Pettit (2005) and Verburg,
Schot, and Veldkamp (2004), we adapted the criteria and clas-
sified some commonly used models into five groups (Table 6)
according to some of the most popular planning tasks: land-use/
land-cover change, urban growth, transportation land use,
impact assessment, and comprehensive projection.
The main objective of the models in the first group of
Table 6 is to understand LUC planning, for example, the
SLUCE project aims to understand the role of complexity in
structuring land-use patterns and their subsequent ecological
effects by building multiple agent-based models. A good exam-
ple of urban growth model is the SLEUTH model, which derives
its name from the six types of data inputs: Slope, land use, urban,
exclusion, transportation and hill-shading. This model simulates
the transition from nonurban to urban land use using grids of
cells (cellular automaton). UrbanSim is also an urban growth
model, which includes more urban market dynamics by imple-
menting a perspective on urban development that represents a
dynamic process. This process results from the interaction of
many actors on making decisions within the urban markets for
land, housing, nonresidential space, and transportation.
UPLAN and CUF are, to a certain extent, similar to
SLEUTH and UrbanSim and fall into the group of urban
growth models as most of their designed modules, scenarios,
and input/output data focus on urban growth issues (how to
facilitate growth). CUF, for example, as an urban growth
model, recognizes four types of land use: (1) undeveloped land,
(2) residential (single family or multifamily), (3) commercial
(retail and offices), and (4) industrial. The urban growth mod-
ule in UPLAN is based on the attractiveness of landscape
features (such as highways on ramps, roads, cities, and public
transportation) as well as on growth constraints (such as exist-
ing development, open space, and steep slopes).
Admittedly, to some extent, there is no clear boundary
between urban growth and land-use models, because some
urban growth models may further include coupled land-use
modules or vice versa. For example, the SLEUTH model is
taken as an urban growth model, but it also includes a module
that can simulate multiple land-cover class transitions (DEL-
TATRON). The CVCA is a landscape model coupled with
SLEUTH to operate a set of landscape ecological strategies for
urban planning (Silva, Wileden, and Ahern 2008). Neverthe-
less, it is reasonable to make a broad classification such as
dividing the models into growth models (those facilitating
growth) and land-use models (those negotiating among differ-
ent possible land uses) in some interrelated research areas just
to illustrate their main functionality/purpose.
Therefore, our classification mainly considers the models’
applications; for example, Land-Use Change Analysis System
(LUCAS—Pullar and Pettit 2003) is taken as a land-use model
in this classification, because it generates new maps of land
cover representing the amount of land-cover change to address
issues such as biodiversity conservation, assessing landscape
elements and long-term landscape integrity. In the transporta-
tion land-use group, ILUTE is a typical integrated model,
which includes four core components: land development, loca-
tion choice, activity/travel, and auto ownership. Each of these
behavioral components involves a complex set of submodels
that incorporate supply/demand interactions and subsequent
interactions among each other. For example, land use evolves
Table 5. Models with Different Spatial or Aspatial Emphasis
Benchmark of Classification: Spatial and Aspatial highlights
Classifications of
Models Models References
Spatial oriented CLUE de la Barra (2001)
NELUP O’Callaghan (1995)
SLEUTH Silva and Clarke (2002)
Aspatial oriented DELTA Simmonds (1999)
EMPIRIC Rothenberg-Pack (1978)
LINE Madsen and Jensen (2000)
MEPLAN Echenique and Hunt (1993)
TRANUS de la Barra (2001)
Integrated models CVCA Silva, Wileden, and Ahern
(2008)
DG-ABC Wu and Silva (Forthcoming)
IMAGE-GTAP/LEI Klijn et al. (2005)
ILUTE Miller (2001)
METROSCOPE Larson, Cser, and Conder
(2000)
Table 6. Models with Different Planning Tasks Emphasis
Benchmark of Classification: Planning Tasks
Classifications of Models Models References
Land-use/land-cover
change
LEAM Deal (2001)
LUCAS Berry et al. (1996)
SIMLAND Wu (1998)
SLUCE Brown (2005)
What if? Klosterman (2001)
Urban growth CUF Landis (2001)
DG-ABC Wu and Silva (Forthcoming)
SLEUTH Silva and Clarke (2002)
UPLAN Walker et al. (2007)
UrbanSim Waddell et al. (2003)
Transportation land use IRPUD Wegener (1998)
ILUTE Miller (2001)
METROSIM Anas (1982)
TLUMIP Weidner et al. (2006)
Impact assessment CommunityViz Kwartler and Bernard
(2001)
INDEX Allen (2001)
Place3S Snyder (2001)
Comprehensive
projection
METROPILUS Putman and Shih-Liang
(2001)
SOLUTION Echenique (2004)
SPARTACUS Lautso (2003)
Silva and Wu 143
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

in response to the location of the needs of households and
firms, and people relocate their homes and/or jobs, at least par-
tially, in response to accessibility factors. TLUMIP is used to
model eight scenarios and combines land use and economic
and transportation policy options.
CommunityViz, INDEX and Place
3
S are rule-based impact
assessment models, that is, the models in this classification nor-
mally do not include procedures for projecting future condi-
tions; instead, they require the user to specify what the future
will be and use clearly defined rules (often incorporated into
a spreadsheet) to measure and evaluate the impacts that an
assumed future state will have on variables of interest (Kloster-
man and Pettit 2005). By specifying rules, they determine and
assess the impacts of alternative public choices and future
development patterns. This classification mainly considers the
models’ primary function/design purpose; therefore, inter-
changes may happen between the categories in Table 6.
For example, although Land-use Evolution and impact
Assessment Model (LEAM; Deal and Pallathucheril 2003,
2009) is listed in the land-use/land-cover change category, it
also includes an impact-assessment module. As Deal and Pal-
lathucheril (2007) pointed out, the LEAM model consists of
two major organizational parts: a LUC model—defined by a
dynamic set of submodel drivers that describe the local causal-
ity of change and enable easy addition and removal of variables
and the ability to play out ‘‘what-if’’ scenarios—and impact
assessment models that facilitate interpretation and analysis
of LUC depending on local interest and applicability. Compre-
hensive projection models propose not only LUCs but also
other variables of interest such as transportation flows, popula-
tion, employment, floor space, and environmental impacts. The
large-scale urban models like METROPILUS, SOLUTION,
and SPARTACUS are typical models that fit this classification.
Based on the above analysis and the respective tables, we
can observe that the differences between the classifications
of urban models are not as obvious as one might imagine. For
example, in some cases, there is no clear distinction between
the models used for urban growth and LUC or transportation
because their applications relate very closely to each other (and
usually imply causality relations among variables). A good
example is the SLEUTH model; in the literature, it is taken
as an urban growth model, but it also includes a module that
can simulate multiple land-cover class transitions (DELTA-
TRON). The LEAM determines the growth potential of all
land; at the same time, it is used to predict urban growth and
the environmental impacts associated with urban growth based
on physical and social factors. Therefore, to some extent, it is
hard to classify LEAM as a land use or urban-growth model.
Undoubtedly, the classification schemes and models listed
are only representative of the mainstream literature (there are
other models developed and not published in the literature) and
are briefly introduced in this article—as the objective of this
part of our article is to give a brief introduction to the different
characteristics of currently used urban models and having as
reference the above classifications.
An Overview of Existent Urban Models and a
‘‘Zoom in’’ to Dynamic Models
Overview of Urban Models
In an attempt to bring all the models together in one unique
synthesis table, we propose Table 7. A list of models is juxta-
posed to the main title/goal of each previous table (Tables 1–6).
We followed the information on the previously mentioned
models in order to fill the different boxes. As is the reader can
see, many of the cells in this table are not filled (pointing to the
need for more metadata provided by the developers of the
models).
The identification of the modeling approach (mostly GIS,
CA, and ABM models), the planning task (mainly Land Use
and Transportation) and the level of analysis (micro, macro,
cross level) tend to be the usual information provided by the
authors. In terms of the main characteristics, CLUE, CVCA,
DELTA, PUMA, GSM, SLEUTH, LTM, DDG-ABC, LEAM,
ILUMASS, CURBA, and SAM-IM seem to include the infor-
mation regarding up to six of the main benchmarks defined for
the analysis of the models, while the majority of the models
will only include three benchmarks (and in many cases, only
two). This reveals an important flaw in the provision of
meta-information for the models.
In conclusion, our classification covers six important
aspects of urban models: modeling approaches, levels of anal-
ysis, spatial and temporal scales, spatial and aspatial dynamics,
and planning tasks. We discovered that the majority of the
models now include multiple disciplinary requests in order to
solve problems of increasing complexity (Wu and Silva
2009). For instance, CVCA (Silva, Wileden, and Ahern
2008) requires the contribution of geographers (regarding
urban growth and land-use change), planners (in order to
explore issues of planning policy and practice—i.e., the policy
goals of establishing greenway corridors and the practicalities
of doing so in urban environments) and landscape ecologists
(as the metrics of patterns and processes require a good under-
standing of size and proximity of forestry/agriculture/parks
patches in order to increase species connectivity). To take
another integrated model as an example, DG-ABC includes
both spatial and aspatial dynamics. Obviously, the analysis and
modeling of these dynamics require an understanding of spatial
issues such as geographical information but also economics
and psychological/social theory and practice. As a conse-
quence, the techniques increasingly support intelligence-led
solutions as a way to include multidisciplinary knowledge and
the integration of different scales and timeframes. As a result,
these new concerns are leading to a new generation in the sys-
tems evolution of urban modeling.
While all previous modeling approaches identified in the
different tables are interesting for the purpose of this article and
as a testimony of the innovative research, it is important to
zoom in and detail the dynamic modeling approaches (as these
are considered as being at the forefront of the new generation of
models), and among those models, concentrate just on the ones
144 Journal of Planning Literature 27(2)
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

Table 7. An Overview of Urban Models
Model Name
Modeling
Approaches
Level of Analysis:
Micro, Macro,
Cross
Spatial Scales: Regional,
Metropolitan, Local,
MultiScale
Temporal Scales:
Long, Medium,
Short
Spatial Context:
Spatial, Aspatial,
Integrated
Planning
Tasks
a
Total
Agent-LUC A M L 3
BabyLOV C C S L 4
CARLOS L I 2
CUF R Mu U 3
CUFM G Mi M U 4
CURBA G Ma M S u5
CLUE/CLUES G C Mu M I L 6
Community Viz R L I 3
CVCA I C M M-L I U 6
DG-ABC I CA M M I u6
DRAM M/S M I 3
DELTA M/S Mi Mu S A T 6
DUEM C Mu u3
EMPIRIC M/S Au3
Environment Explorer C C Mu I L 5
GEOMOD2 G Ma R L 4
GSM G Ma Mu S L 5
INDEX R I2
ILUTE I Mi I T 4
ITLUP I L2
IMAGE-GTAP/LEI G IL3
IRPUD G Ml T3
IIASA Ma L2
ILUMASS G Mi M I T 5
IFDM L T 2
LOV C Ma L3
LEAM C Mi R I I 5
LUSD C Mu L 3
LINE M/S AC3
LUCAS G IL3
LUCIM A L L 3
LTM G Ma S L 4
LUCITA A L2
LTM G R M S L 5
METROPILUS G R C 3
METROSCOPE G L I U 4
METROSIM G Ml I T 4
MEPLAN M/S L A T 4
NELUP SC2
Place3S R L L 3
POL IS M/S Mi U3
PLUM M/S AL3
PUMA A Mi M L I U 6
FEARLUS A Mu L I L 5
PECAS G L S T 4
SAM-IM R R S S U 5
SLEUTH C C Mu S U 5
SPARTACUS G R C 3
SOLUTIONS G Mu T 3
SelfCormas A S S L 4
SYPRIA A L2
STIT M/S T2
SIMLAND I L2
SLUCE A L2
SLUDGE I L2
SIMPOP A Mi U3
TRANUS M/S R A T 4
(continued)
Silva and Wu 145
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

that provide examples of an application into case studies.
Therefore, the next section is entitled ‘‘Analysis of Urban
Dynamic Models’’ and will allow us to perform a further char-
acterization of these models.
Analysis of Urban Dynamic Models
The previous section attempted to present a comprehensive
view of the models presented in the literature during recent
years. Over these last few decades, advances in computation,
in the knowledge of the different systems and accumulated
expertise of researchers working in these areas are some of the
main causes for the development of a new generation of mod-
els, which are dynamic in nature. This section ‘‘zooms in’’ on
the previous classification and details these dynamic models. It
starts by listing and contrasting dynamic models that address
growth and land-use particulars.
Currently, urban models are necessary and important tools
for urban planning. There is much information to help us not
only in structuring the development of this diverse set of tech-
nologies but also to lay the foundations for assessing the advan-
tages and disadvantages of the application of different models
for different tasks in different contexts. In this part of the article,
we target only the dynamic models as previously described.
First, we roughly categorize the models into two groups:
models for urban growth (Table 8) and models for LUC
(Table 9). Urban growth models concentrate on urban expan-
sion. Most of their application areas or analyses are related to
urban regions and the growth phenomena. LUC models mainly
focus on land-use and land-cover changes; for example, the
models that evaluate the transition of different land types for
the purpose of optimizing land configurations (which best
match a set of goals and objectives) or for the purpose of ana-
lyzing the characteristics of different patches/patterns/pro-
cesses of land use in order to create sustainable development.
Second, in each group we analyze some representative
dynamic models with four factors: (1) objectives, (2) spatial
scale, (3) temporal scale, and (4) the information required.
These factors are the four most commonly introduced items
of information in the description of the urban land-use models
and therefore can reflect to a certain extent the main character-
istics of the models allowing for comparisons to be made.
Among these models, SLEUTH, LTM, LUSD, DUEM, and
FEARLUS are CA-based models. A large and increasing vol-
ume of work shows that CA are proper tools for modeling spa-
tial dynamics. CA have a natural affinity to represent complex
spatial forms through relatively simple rules (Silva 2008a;
Silva and Clarke 2005; Wu and Silva 2010; Pinto and Antunes
2007; Mavroudi 2007; Batty 2005, 2007). Take the SLEUTH
model for example; it produces four types of urban growth:
spontaneous, diffusive, road-influenced, and organic growth-
based on transition rules. It simulates the transition from
nonurban to urban land use using a grid of cells (cellular auto-
mation), the state of each of which is dependent upon the rules
and five coefficients (diffusion, breed, spread, slope, and road
gravity) that are calibrated with historical growth data. In con-
trast with CA, the agent-based model has its strongest represen-
tation on aspatial dynamics.
Agents are good at simulating individual decision-making
entities and their interactions to incorporate social processes
on decision making and dynamic socioeconomic environmen-
tal linkages. For example, in the FEARLUS model, the agents
work as land managers to decide each year how to use the land
and how many land parcels they will own. The model uses a
decision algorithm, which is applied to each individually
owned land parcel, rather than to the farm as a whole. SOLU-
TIONS, TRANUS, FEARLUS, and SAM-IM also provide
insights into microscopic simulation including behavioral mar-
ket and spatial policy simulation. In terms of spatial and tem-
poral scales among these models, they vary from a fixed
scale (e.g., UrbanSim, DUEM, LUCAS, LTM, and UPLAN)
to a flexible definition by users (METROSIM, CUF, GSM, and
SOLUTIONS).
Conclusion: In the Fifth Generation and the Future of
Urban Models
The dynamic models of urban growth and LUC have evolved
significantly since their early applications decades ago. When
Table 7. (continued)
Model Name
Modeling
Approaches
Level of Analysis:
Micro, Macro,
Cross
Spatial Scales: Regional,
Metropolitan, Local,
MultiScale
Temporal Scales:
Long, Medium,
Short
Spatial Context:
Spatial, Aspatial,
Integrated
Planning
Tasks
a
Total
TLUMIP M/S Ml T3
UrbanSim I Mi Mu L 4
UPLAN R Mu S U 4
UGM C Mu L S U 5
UED C Ma S U 4
What if? R L2
WiVsim C L T 3
Total ¼64 models 57 24 31 13 29 62
Note: Modeling Approaches: M/S ¼mathematical /statistical models; G ¼GIS-based models; C ¼cellular automata-based models, A ¼agent-based models;
R¼rule-based models; I ¼integrated models.
a
Planning tasks:L¼land-use/land-cover change; U ¼urban growth; T ¼transportation land use; I ¼impact assessment; C ¼comprehensive projection.
146 Journal of Planning Literature 27(2)
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

it comes to assessing and comparing different urban models, it
is essential to consider the purpose for which each is going to
be used and the results that are expected from each one. In
reflecting on the experience of the first-generation models
nearly three decades ago, Batty (1979) noted that models
should be evaluated in terms of their contribution to both sci-
ence and design.
This article builds on previous research (Wu and Silva 2010;
Batty 2007; Issaev et al., 1982; Stahl 1986; Mitasova and Mitas
1998; Briassoulis 2000; Verburg, Schot, and Veldkamp 2004).
It starts by presenting an overview of the main classifications of
the models, building on the previous work of other authors and
clarifying the overall production of models at the present
moment—exploring what the models are, at what scales, for
what purpose the models are built and what the future develop-
ment will be. From the above analysis, we can see that micro
simulation seems to be a popular approach and, among the
modeling approaches, cellular automata, and agent-based mod-
els seem to have gained ground on the number of model appli-
cations, case studies and degrees of operationalization.
Most of the newer generation of urban models are designed
with the objective of making them more policy sensitive.
Unfortunately, most of them have yet to reach a point where
they can be fairly evaluated on this criterion, and the older
operational models still raise important questions about the
utility of such complex tools. It seems that very few models sat-
isfy these requirements with some degree of reliability and
implementability. One conclusion is that already reached by
Benenson and Torrens (2004), Moulin, Chaker, and Gancet
(2004) and Miller et al. (2004) and still far from being refuted:
some models are either highly abstract, missing the policy-
driven context; theoretically interesting but with insufficient
tools for practice; or are in an early stage of their operational
development.
Also, current circumstances are still a long way ahead in
terms of the true integration of theoretical analyses at various
spatial levels and within particular time frames. For instance,
this is the case for the following examples: integration of the
various pieces of knowledge for aggregated modeling (Wege-
ner 2004; White, Straatman, and Engelen. 2004); development
of coherent theories and methodologies to guide future LUC
toward sustainable paths (Briassoulis 2000); and integration
of new theories and techniques (Berling-Wolff and Wu
2004). Even though these articles have been published and
these issues raised, a wide gap yet remains between theoretical
analysis and practical applications and development.
When the information is provided by the model developer/
user, there are still some issues with verification and precision
of modeling results. The validation and evaluation mechan-
isms, the transferability and reusability still present a gap for
scholars to bridge in the future.
Metadata information and input data to the models remain as
a recurring problem; for instance, in the ‘‘key information
required’’ for the urban growth models (in Table 8 and the
land-use models in Table 9), we can discover that models with
similar planning tasks have very different data analysis and
variables for input information. Nevertheless, results from
these models are very often used to compare performance
Table 8. Comparisons of a number of dynamic models on urban growth
Model Objectives Spatial Temporal Key in Formation Required
CUF Analyze the impacts of alternative
regulatory and investment policy
initiatives
Customized lor
user needs
Five years Population trend data, development cost data,
user-defined map layers
CURBA Urban growth, policy simulation and
evaluation
One-hectare
(100 100 m)
grid cells
User defined Population or household growth data
DELIA Assistant for policy makers and
planners
User defined One year
increments
recommended
Fixed-format ASCII files contains location of
households, economic and demographic
scenarios
DUEM (CA) Exploring urban development Raster 30 30 m Dynamic discrete Housing, manufacturing / primary industry
commerce and services.
METROSIM Simulate urban growth and land-use
change
User defined User defined Transportation, land-use, metropolitan growth
SI. "RUTH
(CA
model)
Understand urban growth by
projecting urban expansion.
User defined
(50 m–1 km)
Yearly Slope, land cover, exclusions, urban areas,
transportation, hydrologic (8-bit Gif file of data
layer)
TRANUS Simulate the location of activities in
space, land use, the real estate
market and transportation system
GIS layers Static equilibrium Population, salaries, profits taxes, imports,
households, transport, jobs
UrbanSim Land use and transport planning, public
policy on urban sprawl
GIS vector based
and grid cells
(150 150 m)
Variable, annual,
dynamic
Parcels, business establishments, household data,
environmentally sensitive layers, zones used in
travel modeling, travel impedance
What if? Analyze land suitability and impacts of
policy choices on land-use patterns
and social trends.
ESRIs shapefile Variable: five to
ten years
Slope, soil, species, stream, permissible land-use
conversions, households, regional
employment.
Silva and Wu 147
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

among models or to provide comparative analysis of planning
outcomes in similar case studies.
The demands of more integrated and sophisticated models
create a need to understand the synergy between interdisciplin-
ary knowledge and integrated approaches in order to represent
different phenomena in urban systems. The issues of interdisci-
plinary knowledge with aggregated model development, mod-
eling steps and support for new techniques require a new
generation of urban dynamic models that better address the
multiscale characteristics of urban land systems, that imple-
ment new techniques to quantify neighborhood effects and that
explicitly deal with temporal and spatial dynamics and that
achieve a higher level of integration between disciplinary
approaches and urban land changes. If these requirements are
fulfilled, the models will better support the analysis of urban
land dynamics.
To achieve adequate reliability of simulation results, some
considerations should be added to the model proposal, design,
and assessment procedures: first, to understand the current
urban models, popular modeling approaches, and their charac-
teristics/applications need to be applied. Our work, and this
article, are an attempt to contribute to the progress in this area.
Second, there is a need to understand the dynamics and pro-
cesses of urban LUC in different temporal and spatial scales
and investigate the principal actors of the model, data collec-
tion and decision-making processes. Third, it is necessary not
only to construct the framework of the model but also to trans-
late it into a flexible and portable simulation tool that can be
integrated with the model at various levels/scales/timeframes.
Fourth, the performance of the model, validation tests, calibra-
tion, and the evaluation of model should all be assessed.
Finally, there is a continuous need to integrate new tech-
niques from multiple disciplinary fields and allow for hybrid
modeling. Urban land dynamics’ simulation is an evolving
research field with a primary focus on exploring new model-
ing paradigms and techniques derived from the advances in
Table 9. Comparisons of a Number of Dynamic Models on Land Use
Model Objectives Spiilial Temporal Key Information Required
CLUE-S Modeling the spatial dynamics of small
regional land use
GTS pixel (a few
meters–1 km)
N/A Land-use data derived from remote
sensing image, fine resolution dita
CLUE National and continental level model for
land use
GIS pixel
(7–32 km)
N/A Coarse resolution data, land-use data
derived from census or survey
FEARLUS (ABM
and CA)
Understanding land-use change in microlevel Raster (CA) IDynamic discrete Environmental and land manager
parameters, horizontal size, land
parcel price, neighborhood weighting
range
GSM Population growth and new development
effects on land-use/land-cover nutrient
pollution loads
User defined User defined Arcinfo format data, including land use,
soil, streams, buffer zones,
population, household, preserved
land, environment data
INDEX Measure the characteristics and
performance of land-use plans and urban
designs
User defined Yearly Regional data, zonal data, transport
network data
LUCAS Modeling land-use and environment changes Raster cell
(90 90 m)
5 years–100 years transportation, slope and elevation,
ownership, land cover, population
density, environmental variables
LTM
(CA model)
Land-use change analysis and physical
impacts
Raster cell
(100 100 m)
Variable five to
ten years
increments
Arcinfo format data, previous land use,
road, surface water, public lands,
population
LUSD
(CA model)
Simulate land-use scenario demands Raster (CA) Dynamic discrete Scenarios factors: population, market,
technology, GDP
NELUP Predicts patterns of agriculture and
forestry land use under various scenarios
1-km cell size N/A Meteorological, Parish census, farm,
species, land cover, soil chrematistics
SAM-IM Regional level or a microscopic level
simulation
User defined User defined Land-use type, industrial employment,
military, office employment, existing
land-use information and scenarios
SOLUTIONS The working of two interrelated markets:
the land market and the transport market
for spatial policy.
User defined User defined The number of dwellings and business
floorspace projections per model
zone, employment, population,
households by size for the modelled
area, car ownership, transport
improvements, costs and capacities
UPLAN Land-use evaluation and change analysis
based on general land-use plans, popula-
tion and employment projections
Raster cell
(50 50 m)
Annual fifty years Demographic and land-use factors,
regional general plans.
148 Journal of Planning Literature 27(2)
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

computation for the goal of enriching the spatial and aspatial
analysis of highly complex, dynamic and often nondeterminis-
tic problems. In this light, recent developments in modeling
techniques, for example, the applications of AI techniques
(such as ‘‘Weak AI’’: technology that is able to manipulate pre-
determined rules and apply the rules to reach a well-defined
goal [Bethell 2006]), seem to prove they are good tools for
mimicking complex urban dynamics.
The demands for more integrated and sophisticated models
create a need to understand the synergy between interdisciplin-
ary knowledge and integrated approaches in order to represent
different phenomena in urban systems. This has been acknowl-
edged by the emerging integrated models, and also coincides
with the notion of the fifth generation of modeling systems
(Chau and Chen 2001) and the trend of the incorporation of
AI procedures anticipated by Openshaw (1992).
While there are still many gaps to be bridged, it is undeni-
able that a great deal has been achieved during the past several
decades. It is undeniable that the future will be one of inte-
grated dynamic models and that many of the research projects
developed in labs during the previous decades are starting to
reach (and are being understood by) practitioners, who consider
it one important tool in their portfolio of methodologies for data
analysis, simulation, scenario development, visualization and
communication, among others.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, author-
ship, and/or publication of this article.
References
Abbott, M. B. 1989. ‘‘Review of Recent Developments in Coastal
Modelling.’’ In Hydraulic and Environmental Modelling of
Coastal, Estuarine and River Waters, edited by R. A. Falconer,
P. Goodwin, and R. G. S. Matthew, 3–39. Aldershot: Avebury
Technical.
Allen, E. 2001. ‘‘INDEX: Software for Community Indicators.’’ In
Planning Support Systems: Integrating Geographic Information
Systems, Models and Visualization Tools, edited by R. K. Brail and
R. E. Klosterman, 229–62. Redlands, CA: ESRI.
Anas, A. 1982. Residential Location Markets and Urban Transporta-
tion: Economic Theory, Econometrics, and Policy Analysis with
Discrete Choice Models. NewYork: Academic Press.
Batty, M. 1976. Urban Modelling: Algorithms, Calibrations, Predic-
tions. Cambridge: Cambridge University Press.
———. 1979. ‘‘Progress, Success and Failure in Urban Modelling.’’
Environment & Planning A 11: 863–78.
———. 1995. ‘‘Planning Support Systems and the New Logic of
Computation.’’ Regional Development Dialogue 16: 1–17.
———. 2004. ‘‘Editorial.’’ Environment and Urban Planning B 31:
326–30.
———. 2005. Cities and Complexity: Understanding Cities with Cel-
lular Automata, Agent-Based Models, and Fractals. Cambridge,
MA: MIT Press.
———. 2007. ‘‘Cities and Complexity: Understanding Cities with
Cellular Automata, Agent-Based Models, and Fractals.’’ Journal
of Regional Science 47: 624–7.
Batty, M., and P. Longley. 1994. Fractal Cities: A Geometry of Form
and Function. London, England: Academic Press.
Batty, M., Y. Xie, and Z. Sun. 1999. ‘‘Modelling Urban Dynamics
Through GIS-Based Cellular Automata.’’ Computers, Environ-
ment and Urban System 23: 205–33.
Benenson, I., and P. M. Torrens. 2004. Geosimulation Automata-Based
Modelling of Urban Phenomena. Chichester, Sussex: John Wiley.
Berling-Wolff, S., and J. Wu. 2004. ‘‘Modeling Urban Landscape
Dynamics: A Review.’’ Ecological Research 19: 119–29.
Berry, M., R. Flamm, B. Hazen, and R. Maclntyre. 1996. ‘‘LUCAS: A
System for Modelling Land Use Change.’’ IEEE Computational
Science and Engineering 3: 24–35.
Brotchie, J. F. 1969. ‘‘A General Planning Model.’’ Management Sci-
ence 16: 265–66.
———. 1978. ‘‘A New Approach to Urban Planning.’’ Management
Science 24: 1753–58.
Brown, D. 2001. ‘‘Project SLUCE,’’ http//www.cscs.umich.edu/
research/projects/sluce
Brown, DanielG. 2005. ‘‘Project SLUCE: Understanding Biocomplexity
at the Urban-Rural Interface, National Science Foundation, Under-
standing and Harnessing Complexity in the Environment.’’ Meeting
of the Biocomplexity in the Environment Awardees, Arlington, Vir-
ginia, March 21-23.
Briassoulis, Helen. 2000. ‘‘Analysis of Land Use Change: Theoretical
and Modeling Approaches.’’ In The Web Book of Regional Science.
Regional Research Institute, West Virginia University. http://www.
rri.wvu.edu/WebBook/Briassoulis/contents.htm
Bru
¨cher, W., C. Kessler, M. J. Kerschgens, and A. Ebel. 2000. ‘‘Simu-
lation of Traffic-Induced Air Pollution on Regional to Local
Scales.’’ Atmospheric Environment 34: 4675–81.
Clarke, K. C., and L. J. Gaydos. 1998. ‘‘Loose-Coupling a Cellular
Automaton Model and GIS: Long-Term Urban Growth Prediction
for San Francisco and Washington/Baltimore.’’ International Jour-
nal of Geographical Information Science 12: 699–714.
Chau, K. Wok-wing. 2006. ‘‘A Review on Integration of Artificial
Intelligence into Water Quality Modeling.’’ Marine Pollution Bul-
letin 52: 726–33.
Chau, K. W., and W. Chen. 2001. ‘‘A Fifth Generation Numerical
Modelling System in Coastal Zone.’’ Applied Mathematical Mod-
elling 25: 887–900.
Cunge, J. 1989. ‘‘Review of Recent Developments in River Model-
ling.’’ In Hydraulic and Environmental Modelling of Coastal,
Estuarine and River Waters, edited by R. A. Falconer, P. Goodwin,
and R. G. S. Matthew, 393–404. Aldershot: Avebury Technical.
D’Aquino, P., C. Le Page, F. Bousquet, and A. et Bah. 2003. ‘‘Using
Self-Designed Role-Playing Games and a Multi-Agent System to
Empower a Local Decision-Making Process for Land Use Manage-
ment: The SelfCormas Experiment in Senegal.’’ Journal of Artifi-
cial Societies and Social Simulation 6, http://www.jasss.soc.
surrey.ac.uk/6/3/5.html
Silva and Wu 149
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

Deal, B. 2001. ‘‘Ecological Urban Dynamics: The Convergence of
Spatial Modelling and Sustainability.’’ Building Research and
Information: 29: 381–93.
Deal, B., and V. Pallathucheril. 2003. ‘‘The Land Evolution and Impact
Assessment Model (LEAM): Will It Play in Peoria?’’ Proceedings of
the 8th International Conference on Computers in Urban Planning
and Urban Management, Sendai, Japan, May 27–29.
———. 2007. ‘‘Developing and Using Scenarios.’’ In Engaging the
Future: Forecasts, Scenarios, Plans, and Projects, edited by L.
D. Hopkins and M. A. Zapata, 221–42. Cambridge, MA: Lincoln
Institute for Land Policy.
———. 2009. ‘‘A Use-Driven Approach to Large-Scale Urban Mod-
elling and Planning Support.’’ In Planning Support Systems Best
Practice and New Methods, edited by S. Geertman and J. C. H.
Stillwell, 29–51. Springer ScienceþBusiness, Sendai, Japan.
De la Barra, T. 2001. ‘‘Integrated Land Use and Transport Modeling:
The TRANUS Experience.’’ In Planning Support Systems: Inte-
grating Geographic Information Systems, Models and Visualiza-
tion Tools, edited by R. K. Brail and R. E. Klosterman, 129–56.
Redlands, CA: ESRI.
Echenique, Marcial. 2004. ‘‘Introduction to SOLUTIONS Research
Project.’’ Cambridge Conference, Cambridge, December 15.
Echenique, M., and J. D. Hunt. 1993. ‘‘Experiences in the Application
of the MEPLAN Framework for Land Use and Transport Interac-
tion Modelling.’’ In Proceeding of the 4th National Conference on
the Application of Transportation Planning Methods, Daytona
Beach, Florida, USA.
Ettema, Dick, Kor de Jong, Harry Timmermans, and Aldrik Bakema.
2005. ‘‘PUMA: Multi-Agent Modelling of Urban Systems.’’ In
45th Congress of the European Regional Science Association,
Vrije Universiteit Amsterdam.
Fischer, G., and L. Sun. 2001. ‘‘Model Based Analysis of Future Land-
Use Development in China.’’ Agriculture, Ecosystems & Environ-
ment 85: 163–76.
Guan, Qingfeng, Liming Wang, and Keith C. Clarke. 2005. ‘‘An Artifi-
cial-Neural-Network-Based, Constrained CA Model for Simulating
Urban Growth.’’ Cartography and Geographic Information Science
32: 369–80.
He, Chunyang, Norio Okada, Qiaofeng Zhang, Peijun Shi, and
Jinggang Li. 2008. ‘‘Modelling Dynamic Urban Expansion Pro-
cesses Incorporating a Potential Model with Cellular Automata.’’
Landscape and Urban Planning 86: 79–91.
He, Chunyang, Yaozhong Pan, Peijun Shi, Xiaobing Li, Jin Chen,
Yuechen Li, and Jianggang Li. 2004. ‘‘Developing Land Use Sce-
nario Dynamics Model by the Integration of System Dynamics
Model and Cellular Automata Model.’’ Proceedings of 2004 IEEE
International Geoscience and Remote Sensing Symposium. Beij-
ing, China.
Hoffmann, M., H. Kelley, and T. Evans. 2002. ‘‘Simulating Land-
Cover Change in South-Central Indiana: An Agent-Based Model
of Deforestation and Afforestation.’’ In Complexity and Ecosystem
Management: The Theory and Practice of Multi-Agent Systems,
edited by M. Janssen, 218–27. Cheltenham, England: Edward
Elgar.
Hunt, J. D., and J. E. Abraham. 2003. ‘‘Design and Application of the
PECAS Land Use Modelling System.’’ In Proceeding of 8th
Computers in Urban Planning and Urban Management Confer-
ence, Sendai, Japan.
Issaev, B., P. Nijkamp, P. Rietveld, and F. Snickars, eds. 1982. Multi-
regional Economic Modelling Practice and Prospects. Amster-
dam: North Holland.
Klosterman, R. E. 2001. ‘‘TheWhat if? Planning Support System.’’ In
Planning Support Systems: Integrating Geographic Information
Systems, Models and Visualization Tools, edited by R. K. Brail and
R. E. Klosterman, 263–84. Redlands, CA: ESRI.
Klosterman, R. E., and C. J. Pettit. 2005. ‘‘An Update on Planning
Support Systems.’’ Environment and Planning B: Planning and
Design 32: 477–84.
Klijn, J. A., L. A. E. Vullings, M. van den Berg, H. van Meijl, R. van
Lammeren, T. van Rheenen, A. Veldkamp, P. H. Verburg, H.
Westhoek, and B. Eickhout. 2005. The EURURALIS Study: Tech-
nical Document. Alterra Report 1196, Alterra, Wageningen.
Kwartler, M., and R. N. Bernard. 2001. ‘‘CommunityViz: An Inte-
grated Planning Support System.’’ In Planning Support Systems:
Integrating Geographic Information Systems, Models and Visuali-
zation Tools, edited by R. K. Brail and R. E. Klosterman, 285–308.
Redlands, CA: ESRI.
Landis, J. D. 1994. ‘‘The California Urban Futures Model: A New
Generation of Metropolitan Simulation Models.’’ Environment and
Planning B: Planning and Design 21: 399–422.
———. 2001. ‘‘CUF, CUF II, and CURBA: A Family of Spatially
Explicit Urban Growth and Landuse Policy Simulation Models.’’
In Planning Support Systems: Integrating Geogrphic Information
Systems, Models, and Visualization Tools, edited by R. K. Brail
and R. E. Klosterman, 157–200. Redlands, CA: ESRI.
Larson, Karen, Jim Cser, and Sonny Conder. 2000. ‘‘Metroscope:
Simulating Future Urban Landscapes at the Parcel Level.’’ Twen-
tieth ESRI International User Conference June 26–30, http://www.
oregon.gov/ODOT/TD/TPAU/docs/Symposiums/Second/Conder.
pdf?ga¼t
Lautso, K. 2003. ‘‘The SPARTACUS System for Defining and Ana-
lyzing Sustainable Land Use and Transport Policies.’’ In Planning
Support Systems in Practice, edited by S Geertman and J Stillwell,
453–63. Heidelberg, Germany: Springer.
Lim, K, P. J. E. Deadman, E. Brondizio Moran, and S. McCracken.
2002. ‘‘Agent-Based Simulations of Household Decision-Making
and Land-Use Change Near Altamira, Brazil.’’ In Integrating Geo-
graphic Information Systems and Agent-based Modelling Tech-
niques, edited by H. R. Gimblett, 277–310. Santa Fe Institute
Studies in the Sciences of Complexity. New York: Oxford Univer-
sity Press.
Lee, D. B. 1973. ‘‘Requiem for Large-Scale Models.’’ Journal of the
American Institute of Planners 39: 163–78.
Los, M. 1973. Spatial Design and Artificial Intelligence. Philadelphia:
University of Pennsylvania.
Lowry, I. 1964. A Model of Metropolis. Santa Monica, CA: RAND,
RM-4035-RC.
MAG (The Maricopa Association of Governments), http://www.mag.
maricopa.gov/pdf/cms.resource/subarea_allocation_model32517.
pdf
Madsen, B., and C. N. Jensen-Butler. 2000. LINE: An Interregional
Model for Denmark. Copenhagen, Denmark: AKF.
150 Journal of Planning Literature 27(2)
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

Manson, S. M. 2005. ‘‘Agent-Based Modeling and Genetic Program-
ming for Modelling Land Change in the Southern Yucatan Penin-
sular Region of Mexico.’’ Agriculture, Ecosystems & Environment
111: 47–62.
Mitasova, H., and L. Mitas. 1998. ‘‘Process Modeling and Simulations,
NCGIA GISCC Unit130.’’ http://www.ncgia.ucsb.edu/giscc/units/
u130/u130.html.
Mavroudi, Antonia. 2007. ‘‘Simulating City Growth by Using the Cel-
lular Automata Algorithm.’’ M.Sc. diss., In Adaptive Architecture
& Computation, University of London.
Mensink, C., and G. Cosemans. 2005. ‘‘Determination of Urban Back-
ground Concentrations Caused by Traffic Flows in City Street
Canyons.’’ In Proceedings of the 14th International Scientific Sym-
posium Transport and Air Pollution, Austria, 323–30.
Miller, E. J. 2001. ‘‘Integrated Land Use, Transportation, Environ-
ment (ILUTE) Modelling System,’’ accessed October 9, 2008,
http://www. ilute.com/.
Miller, E. J., J. D. Hunt, J. E. Abraham, and P. A. Salvini. 2004.
‘‘Microsimulating Urban Systems.’’ Computers, Environment and
Urban Systems 28: 9–44.
Moulin, B., W. Chaker, and J. Gancet. 2004. ‘‘PADI-Simul: An
Agent-Based Geosimulation Software Supporting the Design of
Geographic Spaces.’’ Computers, Environment and Urban Systems
28: 387–420.
Nuzzolo, A., and P. Coppola. 2005. ‘‘STIT: A System of Mathemat-
ical Models for the Simulation of Land-Use and Transport Inter-
actions.’’ Proceedings of the European Transport conference,
Strasbourg.
O’Callaghan, J. 1995. ‘‘NELUP: An Introduction.’’ Journal of Envi-
ronmental Planning and Management 38: 5–20.
Openshaw, Stan. 1992. ‘‘Some Suggestions concerning the Develop-
ment of Artificial Intelligence Tools for Spatial Modelling and
Analysis in GIS.’’ The Annals of Regional Science 26: 35–51.
Parker, D. C., and R. I. Najlis. 2003. ‘‘Using Multi-Agent System
Models to Link Spatial Externalities and Landscape Fragmenta-
tion: A ‘Pseudo-Inductive’ Analysis.’’ Paper presented in the
Framing Land Use Dynamics, Utrecht University, Netherlands,
April 16–18.
Pettit, C., J. Keysers, I. Bishop, and R. Klosterman. 2008. ‘‘Applying
the What If? Planning Support System for Better Understanding
Urban Fringe Growth.’’ In Landscape Analysis and Visualisation,
Spatial Models for Natural Resource Management and Planning,
edited by C. Pettit, W. Cartwright, I. Bishop, K. Lowell, D. Puller,
and D. Duncan, 435–54. Berlin, Germany: Springer-Verlag.
Pijanowski, B. C., S. H. Gage, D. T. Long, and W. C. Cooper. 2000.
‘‘A Land Transformation Model: Integrating Policy, Socioeco-
nomics and Environmental Drivers Using a Geographic Informa-
tion System.’’ In Landscape Ecology: A Top Down Approach,
edited by Larry Harris and James Sanderson, 183–98. Bocaraton:
Lewis Publisher.
Pinto, N. N., and A. P. Antunes. 2007. ‘‘Cellular Automata and Urban
Studies: A Literature Survey.’’ ACE: Architecture, City and Envi-
ronment 1: 368–99.
Polhill, J. G., D. C. Parker, and N. M. Gotts. 2005. ‘‘Introducing Land
Markets to an Agent Based Model of Land Use Change: A Design.’’
Paper presented at the Representing Social Reality: Pre-proceedings
of the Third Conference of the European Social Simulation Associ-
ation, Koblenz, Germany.
Pontius, R. G., Jr., J. Cornell, and C. Hall. 2001. ‘‘Modeling the Spatial
Pattern of Land-Use Change with GEOMOD2: Application and
Validation for Costa Rica.’’ Agriculture, Ecosystems & Environ-
ment 85: 191–203.
Power, D. 2004. ‘‘What is a Planning Support System?’’ DSS News
5(12).
Prastacos, P. 1985. ‘‘Urban Development Models for the San Fran-
cisco Region: From PLUM to Polis.’’ Transportation Research
Record 1046: 37–44.
Pullar, D., and C. Pettit. 2003. ‘‘Improving Urban Growth Forecasting
with Cellular Automata: A Case Study for Hervey Bay.’’ In Integra-
tive Modelling of Biophysical, Social and Economic Systems for
Resource Management Solutions, edited by D. Post. Canberra,
Australia: Modelling and Simulation Society of Australia and New
Zealand. http://mssanz.org.au/MODSIM03/Volume_04/C10/02_
Pullar.pdf
Putman, S. H. 1983. Integrated Urban Models. London, England:
Pion.
———. 1995. ‘‘EMPAL and DRAM Location and Land Use Models:
A Technical Overview.’’ In Land Use Modeling Conference Pro-
ceedings, Dallas, TX.
Putman, S. H., and C. Shih-Liang. 2001. ‘‘The METROPILUS Plan-
ning Support System: Urban Models and GIS.’’ In Planning Sup-
port Systems: Integrating Geographic Information Systems,
Models and Visualization Tools, edited by R. K. Brail and R. E.
Klosterman, 99–128. Redlands, CA: ESRI.
Rajan, K. S., and R. Shibasaki. 2001. ‘‘A GIS Based Integrated Land
Use/Cover Change Model to Study Agricultural and Urban Land
Use Changes.’’ Paper presented at the 22nd Asian Conference on
Remote Sensing, Singapore, November 5–9.
Rothenberg-Pack, J. 1978. ‘‘Urban Models: Diffusion and Policy
Application.’’ Monograph Series No.7. Philadelphia, PA: Regional
Science Research Institute.
Sanders, L., D. Pumain, H. Mathian, F. Gue´rin-Pace, and S. Bura.
1997. ‘‘SIMPOP: A Multiagent System for the Study of Urban-
ism.’’ Environment and Planning B: Planning and Design 24:
287–305.
Silva, Elisabete A. 2008a. ‘‘Complexity and CA, and Application to
Metropolitan Areas.’’ In A Planners Meeting with Complexity, edi-
ted by G. D. Roo and Elisabete A. Silva, 309–31. Aldershot, Eng-
land: Ashgate.
———. 2008b. ‘‘Waves of Complexity. Theory, Models, and Prac-
tice.’’ In A Planners Meeting with Complexity, edited by G. D. Roo
and Elisabete A. Silva. Aldershot, England: Ashgate.
Silva, E. A., and K. C. Clarke. 2002. ‘‘Calibration of the SLEUTH
Urban Growth Model for Lisbon and Porto.’’ Computer, Environ-
ment and urban systems 26: 525–52.
Silva, Elisabete A., and Keith Clarke. 2005. ‘‘Complexity, Emergence
and Cellular Urban Models: Lessons Learned from Appling
SLEUTH to Two Portuguese Cities.’’ European Planning Studies
13: 93–115.
Silva, E. A., J. Wileden, and J. Ahern. 2008. ‘‘Strategies for Landscape
Ecology in Metropolitan Planning: Applications using Cellular
Automata Models.’’ Progress in Planning 70: 133–77.
Silva and Wu 151
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from

Simmonds, D. C. 1999. ‘‘The Design of the DELTA Land-Use Mod-
elling Package.’’ Environment and Planning B 26: 665–84.
Snyder, K. 2001. ‘‘Tools for Community Design and Decision-Mak-
ing.’’ In Planning Support Systems in Practice, edited by S. Geert-
man and J. Stillwell, 99–120. Heidelberg, Germany: Springer.
Spahn, M., and B. Lenz. 2007. ‘‘Application of a Multi-Level Freight
Transport Demand Model to Estimate the Effects of Political Mea-
sures.’’ In Young Researchers Seminar 2007, Brno, Czech Repub-
lic, May 27–30.
Stahl, K. 1987. ‘‘Theories of Urban Business Location.’’ In Handbook
of Regional and Urban Economics. Vol. 2, edited by E. S. Mills,
759–820. Amsterdam, North-Holland.
Tobler, W. R. 1979. ‘‘Cellular Geography.’’ In Philosophy in Geogra-
phy, edited by S. Gale and G. Olsson, 379–86. Reidel Publishing
Company, Dordrecht.
Voorhees, A. M. 1959. ‘‘Land Use and Traffic Models.’’ Journal of
the American Institute of Planners 25: 55–57.
Verburg, P., M. Dijst Schot, and A. Veldkamp. 2004. ‘‘Land Use
Change Modelling: Current Practice And Research Priorities.’’
Geojournal 61: 309–24.
Verburg, P. H., H. G. J. de Koning, K. Kok, A. Veldkamp, and J.
Priess. 2001. ‘‘The CLUE Modelling Framework: An Integrated
Model for the Analysis of Land Use Change.’’ In Land Use and
Cover Change, edited by R. B. Singh, Jefferson Fox, and Yukio
Himiyama. Enfield, NH: Science Publishers.
Vonk, G., S. Geertman, and P. Schot. 2005. ‘‘Bottlenecks Blocking the
Widespread Use of Planning Support Systems.’’ Environment and
Planning A 37: 909–24.
Waddell, P., Alan Borning, Michael Noth, Nathan Freier, Michael
Becke, and Gudmundur Ulfarsson. 2003. ‘‘Microsimulation of
Urban Development and Location Choices: Design and Implemen-
tation of UrbanSim.’’ Preprint of an article that appeared in Net-
works and Spatial Economics 3: 43–67.
Wagner, P., and M. Wegener. 2007. ‘‘Urban Land Use, Transport and
Environment Models: Experiences with an Integrated Mircro-
scopic Approach.’’ Dis P 170 (3/2000) 45–56.
Walker, W. Thomas, Shengyi Gao, and Robert A. Johnston. 2007.
‘‘UPlan: Geographic Information System as Framework for Inte-
grated Land Use Planning Model.’’ Transportation Research
Record 1994: 117–27.
Wegener, M. 1998. ‘‘The IRPUD Model: Overview,’’ http://irpud.
raumplanung.unidortmund.de/irpudpro/mod/mod.htm
———. 2004. ‘‘Overview of Land Use Transport Models.’’ In Hand-
book in Transport, Vol. 5. Transport, Geography and Spatial
Systems edited by D. Hensher, K. J. Button, K. E. Haynes, and
P. R. Stopher, 127–46. Oxford: Pergamon/Elsevier.
Weidner, T., R. Donnelly, J. Freedman, J. E. Abraham, and J. D. Hunt.
2006. ‘‘TLUMIP-Transport Land Use Model in Portland -Current
State.’’ In Stadt Region Land 81, 91–102. Aachen: Institut fu
¨r
Stadtbauwesen und Stadtverkehr, RWTH Aachen.
White, R., and G. Engelen. 2000. ‘‘High-Resolution Integrated Mod-
eling of the Spatial Dynamics of Urban and Regional Systems.’’
Computers, Environment and Urban Systems 24: 383–400.
White, R., B. Straatman, and G. Engelen. 2004. ‘‘Planning Scenario
Visualization and Assessment—A Cellular Automata Based Inte-
grated Spatial Decision Support System.’’ In Spatially Integrated
Social Science, edited by M. F. Goodchild, D. G. Janelle, and
Z. G. Shrore, 420–42. Oxford University Press.
Wu, F. 1998. ‘‘SimLand: A Prototype to Simulate Land Conversion
Through the Integrated GIS and CA with AHP-Derived Transition
Rules.’’ International.Journal of Geographical Information Sci-
ence 12: 63–82.
Wu, N. and E. A. Silva. 2010. ‘‘Artificial Intelligence Solutions for
Urban Land Dynamics: A Review.’’ Journal of Planning Litera-
ture 24: 246–65.
———. Forthcoming. ‘‘DG-ABC: An Integrated Multi-Agents and
Cellular Automata Urban Growth Model.’’ Journal of Computer,
Environment and Urban System.
Bios
Elisabete Silva is University Lecturer (Assistant Professor) at the
Department of Land Economy, University of Cambridge. Her research
interests are centered on the application of new technologies to spatial
planning in particular city and metropolitan dynamic modeling
through time. In particular using the following subject areas: Land
Use, Transportation and Metropolitan Planning; Regional Planning,
Integrated Planning; GIS and PPS, and AI Models (CA/ABM/GA).
Recent books published by Dr. Silva, include: ‘‘Urban Revitalization
in traditional neighbourhoods in China’’ (together with Wu and Song)
RICS, 2010; and: ‘‘A Planner’s encounter with complexity’’ (together
with dr Roo), Ashgate, 2010.
Ning Wu is the Pollman Fellow of Harvard Real Estate Academic Ini-
tiative, Harvard University. Before working at Harvard, Ning Wu was
a PhD student at Department of Land Economy, Cambridge
University.
152 Journal of Planning Literature 27(2)
at CAMBRIDGE UNIV LIBRARY on June 4, 2012jpl.sagepub.comDownloaded from