the iconic models that had dominated physical planning hitherto. In this paper, we ... to develop good theory and applications in countless .... MIT's Media Lab.
JAG
Models in planning: changing roles Michael
KEYWORDS:
Spatial Analysis, University College London,
definitions,
computer
transportation,
models,
public
urban
urban
models,
form,
1-19 Torrington
land
Before
symbolically
GIS, 3-dimensional
complex
data sets could be routinely
systems suddenly
ever,
begin
the
idea of a ‘model’
Models are everywhere.
has the drawing
power
architectural
changed.
Before
representations
models domrnated
It once did.
came to dominate
the
focus is once more on traditional models of three-dimensional
1980s and
the
1990s.
representatrons
form,
form.
hitherto.
of modelling
and suggest that new kinds of
as any is contained
Herbert
Simon of
[I 9821, and finally Models
WHAT
ARE MODELS?
ture
of
[I9211
reality”
WHAT
[Paragraph
Logico-Philosophicus]
“A model
2.12,
suggests
page
that
the
model
is rooted
in the philosophy
entific
method.
Yet it was not until
idea of a model then
began
cation’
rather
than
to
develop
good
into fashion was
computing be actually tions
built
of ‘model’
of reality,
as well
as prototypical
of reality’
pace as a vehicle
and
applications
and social sciences. America into
exploration,
the
19651.
where
various and
and
operated designs
large
business.
the notion
better
faith
to
that
make
Echenique
came
‘model’
projects
in
model
in
resentation
252
itself
in an article
used In fact
in a technibased.
characteristic
some 30 years ago, says: “A of a reality,
of the observed
as well
in the
using
is being
language.
its usage was broadly
is made by the expression
[page
clearly
through
are being written
term
or
applications
example
professional
for definition
is a representation
the reality
the
[I 9721 for example,
or may exist”
systems.
books
planning,
ity consists of objects
predic-
of complex
has come to
it first came to be used generally
characteristics
could
while
in everyday
of our concern
Advances better
for a variety
title
for
all
of ‘simplifications’
can be seen quite
planning
and fewer
cal sense within
in science
models
This trend
regional
even when
on which
in countless The term
and
increasingly
to its use in the 1960s as
gathered theory
[Lowry,
inventions.
to hold
itself
‘pictures’
fact that fewer
was as ‘a simplifi-
the term range
urban
being process
to the most routine. its power
possible
and
of
his autobiography
the widest
the
too accelerated
and even
and in sci-
it has lessened
and
translated
space
of science
before
while
Discovery
19891, Models
describe
in science,
first in North
officially
defence,
Tractatus
successively of
widespread,
from the most abstract
as its use has broadened,
1950s that
ascribed
areas of the physical
8,
basic idea of a
were
types of human
Indeed
the
‘a picture
the idea of a model
is a pic-
to be used widely
the classic definition
This was the meaning
or operation,
IS MODELLING?
early definition
in the
of My Life.
has become
many different
a symbol
economist-psy-
Models
[ 1979 and
in 1991 entitled
used to describe
books
[I 9571,
of Thought
Rationality
published
whose
Man
Since the 197Os, the term
Wittgenstein’s
in busi-
of these times
Models
or
edge of quan-
systems theory
of the imagery
Bounded
synthesis are now possible.
the cutting
of the eminent
be
designed
Indeed as good
[ 19771, Models
in urban
thence
and
might
and government.
called:
in but
human
systems
ness, defence,
chologist
In
and technology in the
and
social science and complex
works
about seemed
less glamorous
complex
came to symbolise
did, how-
as scientists
more formally
especially
controlled
ver-
simplifica-
years, the word
of science
domains,
titles of the various
a move back to
physical planning
development
research during the last half century,
But now the
of cities as digital
thus heralding
the iconic models that had dominated we review
physical
its
the 1960s and 1970s. data mod-
lexicon’
understanding
The idea that
hence
restructured titative
1950, models of cities
of
complex
down’
of course
to think
power
for
sciences.
modelled,
However for city systems, the role of mod-
els has been dramatically Mathematical
policy
have pervaded every corner of our world, the
no longer
up the tools
equally
as it was to any funda-
were
‘academic
began
was used for its
‘scaled
sense. The term
systems. In the post-war
providing
of complex systems in science or
which physical
to enter
model
reflected
but in a literal,
to conjure
came of age.
reaching the point where large
manipulated
mental shift in our understanding society. As computers
the term
which
tions
complex
entered the lexicon in the 1960s when the idea of
representing
paper,
purpose
sions of the real thing
participation
This was as much due to computers
this
the mid-century,
and social scientists
The term ‘model’
els then
and
Place, London WC1 E 6BT. UK (e-mail: m.battyQucl.ac.uk)
traditional
ABSTRACT
meant
imperatives
- Issue 3 - 2001
3
Batty
Centre for Advanced
use,
technological
Volume
l
reality
in which of certain and where
or systems that exist, 1641. His focus
as the model
relevant the real-
have existed,
on what
broadens
the rep-
constitutes
this to include
Models planning
past
JAG
as well
as future
unknowable,
al to the fictional, In this paper, history
from
before
Wegener, in terms how
[Harris,
attitudes
respect
to their
themes
that
changing involved thinking
practices the of
fourth
the
involves
their
tools
tangible
to inform
have
ones.
while
what
a
where
the
with
organise
beginning
with
first
this
and inter-
a brief summary
introduced
high
point
that
by then,
their
into planning
was reached
as they
MIT’s
practical
duced.
application
have
Media
We will then
itself,
recount
picking
up on the
new
developments
spurred
in analysis
idea of abstract
historical
perspective
functioning
had
revolution
has continued,
been
earlier
representing
thus forcing
in the physical
of abstraction to the
from
model those
for the emphasis pluralism
also
now
dominates
much expanded
intro-
was rooted
various kinds
to a wider
into
of cities
is more on pedagogic the field.
as models
The planner’s
emphasising
researchers at go back to planning
can
years. As noted,
models’
the
of cities and their
to developments
in systems theo-
economics,
social
physics
and
years, but the shift in thinking such
in these
models
years were
of transportation to land
applications
through
in the
the
followed
ments
its
did
not
location
are
with
and these urban
wave
of
programme
physics
net-
geoning early
separate
land
but
began
very
nevertheless
about
of
regional
were
invented,
thetr
Almost prime
Cities and their
transport
dates for these
new technologies.
box is as
253
a
activities as
by the applicaproblems
which
- social
up in the provided element
as soon as digital
tool
ele-
economics
burthe relat-
computers
focus was to provide
for intensive
computation
essential
from
The second
use. A new
high-
those years.
how urban
to be tied
science
foundations.
studies
had fashioned
complemented
came
was
First for more than
of researchers
but were
ed to computation.
1960s
representation
during
possible.
theories
- all of which field
195Os,
was that
use-transport
in cities. These ideas emerged
intellectual
quickly
In the
planner
tion of classical physics to geographical
to
the begin-
were
of the massive federal
that
a succession
theory
in North
of land use’, and this message
series of rudimentary located
associated
of the
made such modelling
50 years,
emerged
its prediction.
in the wake
Two completely is
New kinds which
dominate
which
studies,
use and
is a function
way building
realisti-
developed,
the field
America
to
and better
of representation
are being
marked
mathematical
which
At the same trme, new kinds of
that
The first
‘traffic
for
link the digital
public
by
models
post-war
presages
but rarely
to reconsider
interfaces
which
one of the watchwords
more
of cities. now
called
in usage in urban
within
revolution
and currently
again
but
until the 1950s.
extended
As the computer
rapidly
cities
nings
and less ambitious
which
up these
but
begin
in a concern
systems
once
through
communications.
mathematical different
computer
representation
have emerged
opening
professionals worked
the field
real world
only
were argue
been
the quest for better
fast toward
roots
had
assumed.
materials
of
years. The
in representation,
has grown
cally,
about
have dominated
culminating
representation
moving
not
the
data for more pragmatic
ends than digital
way
ethereal
versions
toys - the ‘toys of
iconic
‘mathematical
mathematics
way,
how the field withdrew
and simulation,
manipulating
since
represent-
scaled-down
been
Lab. Traditional
our
1960s and we will
ever
imagination.
from traditional
but the change
ry,
most of the key ideas that
at the heart
but with
of the same - digital
to the immediate
in the post-war
in the
visual
public’s
materials
we see manufactured
be traced
of the way models
from
is upon
of the city being
tangible
prehistory
in a straightforward
produced
(http:Nwww.media.mit.eduAoysAotweb/index.html)
in planning.
paper
used in architec-
emphasis
Our icons are no longer versions
within
is that we have now come full
much else in the inter-war We will
‘model’
image of a scale model
at least in the
using models
a balanced
term
and traditionally
the same features
tomorrow’
use, the role of abstract
provide
the
extensively
one of our themes
digital
in
design,
mainstream,
ed no longer
involves
TRANSPORTA-
In a way, this usage still remains
circle with
in
already,
up the popular
materials
Indeed,
USE AND
environment,
and urban
of the
more
be involved
concerns
In con-
together,
UP TO THE 1960s
conjures
appearance.
not only plan-
might
their
- model
role of models
- the
even ‘theory’
to which
themes
and some speculation.
implied
urban
four
of abstraction,
and clients
so that we might
on the wider
planning
exploit we
have
of the ture
have become
thinking, data as models, and participation - involve ideas that we will weave into analysis
in and
with
theme
the field.
all these
LAND
MODELLING
As we
50 years ago was rarely
the extent
ests These four themes
IN THE BEGINNING:
the way those
task. A third
characterise
draw
a synthesis
TION
‘models
-
briefly
1994;
change
model
perhaps
any form
publics
abstract
to
involves
in planning
we will
many
of them,
One of these
term
‘abstractions’,
as being
theme
ners but
field.
role of data which
about
clusion,
this history
of what
continue
the planning
the changing thought
well as process and design
Volume 3 - issue 3 - 2001
the
Batty,
expect
but a second theme
about
to recount
use. In this sense, we will
with
is
attempting
1979;
perceptions
in using models
comfortable
Batty,
we might
characterise
noted
future
the factu-
19981 but to review
significance
already
using
1968;
what
and
from
for this has been done
of the changing are for,
as the
is not merely
modelling
1994; Wilson,
planning’
and extends
science to design.
our concern
of urban
times
systems,
this use of models
l
vehicles
In science and then commerce. systems
provided
prime
candi-
Models planning
The final ming
ingredient
in the
growing and
JAG
that
set the world
1960s was the
and restructuring
more
renewal
mobile,
took
policy
as populations
while
problems
on a new urgency.
of the answer
to getting
stituted
were
Several
because computation were
adversely
conceived
but the singly
were
[Brewer,
were
built
understanding
were
unintelligible
produced
Lee [I9731
a fairly
such complexity.
zones formed
This immediately
of differdecade.
morassed
often
summed
for Large-Scale
All the
ed with
of data
changing solved. what
while
remain
making
the
wanted
was being
seemed
doomed
models
were
mirroring
our
directly
was their
Many
our
models
developers
ceived
of as being
attempted ed with
or its generalisation
to such spatial
how
between
transportation
modelling
ed to reproduce comparative of what look model
this
of a future that were
most often
and
if their
would
even
simply
um. After
return
with
held the key
in the
sectors.
interaction
els [Hill, which
dynamics
and thence
populations
handled
in a more models
crete
time
from
those
such as Lowry’s
models
not on simulating
of
1960s.
from
these
and has influenced obvious
models
in wider structo be
to operating
structure They
over dis-
is little
are
still
different static
and
the fact that
they
many sectors and many other 1974;
was
to embrace
demographic
notwithstanding
[Wilson,
The second of
the
model
Developments discrete
if any
during
that
attempted
to
Batty,
still
modelled,
kinds of
19761.
pretty
spatial
largely
equilibri-
spatial
Pittsburgh
the existing
non-linear model.
resents
city system
within
urban ning
but on opti-
254
wider
but their
a synthesis
misation.
based
space
and the
views
of the
Such models and
based on
have
in mirroring
their
aggregate
real
cousins
19851. The third type of model
but through given
both
the
the notion
had
always
the policy
idea of providing
as
structures
performance
between
Optimisation
modelling
and time
behaviour
model
than
pref-
although
and thus have consider-
of spatial
has been worse
disaggregative
explicitly
Models
tuning
[Ben Akiva & Lerman,
structures
interaction
to treat
process,
disequilibrium.
microsimulation,
one class of mod-
to this. Such models,
are able
finer
the idea of
to economic
city as a system in perpetual much
individuals.
around
linked
contemporary
been embedded
to get
modelling
disaggregation into
for supporting
allow
such as the EMPlRlC in which formed
static,
of the choice
true
cities
is strongly
are central
able potential
much the same for
involved
defining
in transport
theory
attributes
these
of model units
choice which
erence
for it was assumed
were
to those
type
spatial
produced
in kind.
was implicit
[I9641
but their
flow
devel-
way. As part of this, some
intervals in form,
to
later
and employments
generalised
the
it is important
The most
and
integrated
were
in
We will
came to be nested
enabled
issues
resided
emerged
of these
tures which
such
policy
worked.
complements degrees.
always
which
to address
approaches
These models
of these
effort
under
conditioned
economic
spatial
used in
at one level, they just seemed
to
form of cities.
this early
experience
aspatial
state of the city might
the city to a static,
19651 in contrast
attempted
which
in and entrust-
the future
actually
and
behaviour spatial
Thus they
but the real critique cities
extension
can now deal with
in social physics and
change
no real differences
Linear econometric
how
to different
aggregative
for-
in or at least near equilibrium
all, cities had looked
100 years or more; bigger
Few
of urban
that cities were thus
in time.
to make one shot predictions
assumption.
the dynamics
models
in space based on dis-
situation,
static fashion,
made
abstract.
- then models could be calibrat-
like. All the models
time
static,
in the way they
interested
point,
downfall
of
each of which
spatial
- land uses - locat-
concept
the existing
the equilibrium
other
to
con-
If one could develop
activities
aggregative,
the conditions
Three distinct
involved
been
cities were
as accessibility
- the key organising
physical
on the ground
Their failure
this early
opments.
are
as now
at a cross section
interdependence.
mal relationships tance
then
and thus these
how activities
respect to one another
Distance
issue has all but
in equilibrium
to simulate
into
from
to this later but for the moment
note
models, they
vantage
built.
ignorance
return
and
in their
Models’.
In essence,
but rather
about
represented.
away
were
from those
at
and traffic
of abstraction
simplistic
to structures
given
such
to this day although
modelled.
cities were
models
models
to loca-
operated
census tracts
in focus,
decisions
on which
fanciful
real issue related
optimising
a level
these
behaviour,
From our current
that the
it all up in his famous
the computation
However
in that
these
remote
the others Problems
took
hard to relate
made them
behaviourally
was poorly
were
but their
urban
were
economic
introduced
that
‘planning’
more
19721. All these models
and spatial
treated
in
linked level
In short,
physical
incomplete
policy-makers crude,
design.
predetermined
as were
the level at which
the process
Data con-
constraints,
the theory
were
aggregate
A flurry
problems
inevitably
which
like part
and problematic.
budget
to anyone
‘Requiem
[Batty,
of how cities worked.
outputs
implications.
tion decisions
became
to what
was
and
of such efforts
19731. And in parallel,
models
deprivation
remained
biggest
not attuned
poor
paper
by
management
to various
also attempted which
in many ways.
was expensive
it according were
based models
in the subsequent
models
affected
organisational models
made
and many efforts
collection.
goals
in the late 1950s and sev-
was over ambitious
a problem
data
mising
were richer
looked
known.
styles emerged
eral applications The effort
of
hum-
Cities became
Models
to grips with
The story of these years is well ent modelling
of planning
context.
Volume 3 - Issue 3 - 2001
l
and
of economic
opti-
been
context some
rep-
aggregative a theme of urban
unification
in
planto the
JAG
Models planning
field
by linking
they might
the way activities
best locate,
senses, this unification particular
through
elling with ity
emerged
located
to how
quest.
In many
has actually
the linking
discrete
maximising
actually
was an important
from
of spatial
choice theory
although these
been accomplished
few
interaction
tical
in
mod-
[Wilson
models
efforts
et al, 19811.
is particularly
approaches highly
two
these
serious
years,
cities were
conceptualised
efforts
have
of
has now
efforts
quite
19911.
Then
which
there
has been
America
[Echenrque,
tinued
through
than
both
individual gate
tems
There
model
which
work
years
or
without
such
significant
had
we will
[I 9941 provides
a very
world-wide
which
vides a useful paper less
this experience models
such as
with
cities. model
will
in fact
Wegener
is. His recent
of cause
subsequent
commentaries
waves of urban
[see for example, Klosterman
many
the papers
modelling by Batty,
& Lee in the special
change
groups
such ideas
account
for
example. cations
In growth
change
to take place, were
Models
OF
on the
conceived
up until the 1990s
a major
255
being
itself.
radically
The idea of
notions
that
models
element
those
as to how and
radical
cities’
allowed
surprising
ecological
models were
designed
the processes which
by
physics,
in the same way that noise and variation
novel scales
paradigm,
non-equilibrium
for
brfur-
or
at different
Interaction
of the 1960s were
within
of cat-
during
of ‘edge
developed
through
This was as much because
of
linear
But in all these cases, the models
to be operational
predecessor
Harris, Wegener,
ideas
than simple
spontaneous
enablrng
ideas
field),
to speculate
ideas from
[I9921
modelling
dynamics
the agenda
119811 in the spatial
and by Dendrinos
frrst and
issue of the /ourna/
paths,
through
shift.
iconoclastic
the
based on structures
The
fundamental
were
In cities such as the growth
chal-
process
through
began
of the newer,
BUT A REAWAKENING
effect
to
and
rules of urban
and mathematics
now
describe
being
time
more
on
the rnabil-
was
energised
and
in smooth
might
based
by the time
choice theory.
in ways other
years and various
chaos theory. been
on
to deal
such models
his systems
considerably time
models
with
and chaos came onto
change
INTEREST have
effects
[ 19691 somewhat
through
in science
astrophe
to deal
the ground
how to treat
by Wilson
There
In
of oper-
In particular,
a somewhat
to change
discontinuities
20011.
DIRECTIONS
more
unable
evident
behavrour
Jay Forrester’s
space to time
reviewed
are
static
of discrete
models
required
from
Allen [I9971 DIVERGENT
equity,
effects.
ignoring
and to orient
economic
of these
progressions
10 years old, still pro-
we
explicit
failure
about
New
sectoral
by the development
(which
of applications
in the light
that
spatial
was already
lenged
from
1990s.
aggregate
detail
attempt
for
with
his critique.
existing
later.
the action
of
housing
are intrinsically
spatial’,
ity to provide
at Harvard,
at Dortmund
other
of spatral
as well as being
Lee [ 1973) presented deal
but these are one-
summary
nearly
updates
land use and transport
aggre-
early
model
to these
of where
operational
(Wegener,
off
the
to
return
although
measure
efforts
[ZOO01 UrbanSim good
The disillusion
for particu-
although by
range
transport
in time.
them.
however
gone
which
ghetto-rsation,
unambiguous
‘too
rather
and some
Kain
models
have con-
Apart
others,
were
competition
the
applications
and
with
South
sys-
have continued
as Waddell’s
models
glob-
in these
the first and later generations
sectors and groups
from
such as Wegener’s
at Buffalo
efforts
and
with
in the to post-
has become
not on questions
to associate if anything,
other
shortcomings,
spun
at Leeds,
amongst
more,
efforts
and Anas’s efforts
by Wilson
difficult
doubt
on a narrow
and the like which
analysis
homoge-
- urban growth,
market
and
19891 but the model
have
focused
and renaissance,
- but
being interac-
industrial
of the
redevelopment
a
the last
are still
are various
were
The way
and explored
Most
infrastructure
nurtured
all these
in doubt.
to
ational
are not useful
[De La Barra,
at CSIRO,
Individual off
contexts.
developed
Brotchie
from
they
operational
model
they
also
concerns,
into
grave
from
could be tested
issues from a policy standpoint
of Cambridge
disaggregation
simulation,
and suffer
TRAM-S
Echenique
20
have embraced
became
MEPLAN
that
into
at the
companies
environments
is not to say that
lar problem the
consulting
(like
short,
in Europe
organised
of
urban
in equilibrium,
and as the economy
was
1983,
19941. Both these efforts
behavioural
and static
which
applied
spin-off
in the university
Although
Lowry-based
over
society
models
[Putman,
at the University
widely
industnal
within
of movement,
have been thrown
from
model
applied
America
areas,
processes
logis-
failure
relevance.
as being
and as being
al. The way policies
field.
be traced
Pittsburgh
widely
is the
by Echenique
in the
of Putnam can
early
30 years to many cities in North developed
individuals
to work
which
Lowry’s been
many
the
Pennsylvanra
modification
continued
and location,
the
lay in deeper
and of policy
simple
although
for
institutionalised
last 30 years as we have moved
are really
which
continued
DRAM-EMPAL
of
there
around
neous
1960s are still
19941. All of these
the real difficulties of theory
3 - Issue 3 - 2001
that
responsrble
become
in questions
tion
these
at least within
world-wide
notwithstanding
are the
University which
practical
of
and are a part of a new
modelling,
In fact,
who
many
in 1950s and
40 years later,
States.
author)
These
is that
in land use-transport
United
only the
first developed
significant
impetus the
important
largely
to
policy-making,
have
Association,
or lesser extent,
issues were
focused What
~/a~n;~g
to a greater
these
in terms of entropy-utiloperational
insights
the American argue,
Volume
l
and not their
to be.
constitutes such models
Models planning
simulate
and hence
focus.
main
JAG
In
deterministic
the
last
occurred
as the focus
physics,
in
processes with
and
simulated. generate
properties systems urban
outcomes,
critical
model
are currently
these have operational of the
literature
reveals
[see
Portugali,
levels.
content
associated for
on the
essence
few
of
Batty
kinds
&
is close in spirit
with
dynamics
urban are
parallel
and to notions
about
They
1970s the demand
did decline tematic
although
frameworks
interactions
incorporate
growth
is simulated
lution
growth.
once again.
applications.
urban
are once
application
Dortmund
Europe,
models through
are being
various
output
processes.
and
In North America,
as Landis’s
[I9941
do
Waddell’s
through
regional
[Schock,
fall within
Model
science
and
2000].
other
traditions
different
focus on urban
sentation rather
emerging
than
than
pleteness
to
urban
economic
more
to
physical
process,
structure.
to our discussion,
To provide
invoke
are
termed
better with
that
any model
must track
to most of the models cant
exceptions,
Carolina ment that
Chapin
Weiss, and
where
process
built.
in particular the emphasis as a driver
and
19681 were
in this
such interactions
then
But there
Chapin’s
satellite
was on the
of urban
growth.
his colleagues concerned
sense focused
proposed
with
upon
the
the
imagery
have
not
more
traditional
that
as there
driven
by
are explicitthese are
of various
tional
con-
in this newer where
growth com-
traffic
and
signifi-
a superb
in the
models
and there
backcloth
depth
and untried
Nevertheless,
cells are
development,
computed
GIS packages
such as ArcView
lates
growth
(and
starting
in time. to generate thus
decline)
with
urban
In Figure growth
from
development
1, we show consistent
illustrating
how
the
In Figure 2, we show various
is that
spatial
It takes
data
and it simua set of seed between
how with
it
and
interaction
although
explicitly.
desktop
form
256
of
not
over the
diffusion
as differential
from
between
took
in Michigan
sprawl
as well
work.
&
urban
capacity
straints,
process
are developing
Arbor
involving
able
The models
illustrat-
one such model
mechanisms
around
in
for opera-
it is worth
of Ann
with
dangers
models
Batty et a/ [ 1999al town
for a
associated
are clear
makes use of various constraints
with
But such models
of this model
points
land develop-
growth
associated
sprawl.
has been substantial
sites consistent
in North
[Chapin
data
raster based data such as
developed
for the
models.
last 20 years. One of the features
fields
was central were
‘grid-space’
and
provides
tradition.
there
their
incidental.
is that they are closely con-
of urban
purposes.
such a model
much
traditional In fact,
ing the state of the art at this time with
are
are
their
is largely
or
functionality
such untested
policy
models
ideas
‘cell-space’
which
been
developing
a rather
growth
way
urban
the strict limits of CA, and hence they
study of the dynamics
these.
effort
that
and being
than
GIS, particularly
such
some
the idea that land use generated
mor-
processes
with various
interaction.
automata
They rarely
interactions In the 1960s
in
growth
on spatial
cellular
contemporary
to repre-
we need to identify
any
presence
urban
one
modelling
geared
being
in cities,
in such models,
structured
based
to
(CUFj,
there
these
the
transportation
growth,
ways,
main advantages
Most of
seek to provide
contain
driving
in the
sistent
established
which
both
One of their
the traditions
back in the 1950s and 1960s but interestingly,
These
are
dynamics
interaction
and
evaluation
and a series of variants
build on these are being applied these types of model
being
efforts
Futures
not
spatial
pragmatically
appeal
in terms of
to
modelling
Urban
[ZOOO] UrbanSim,
link
20001. that
as a process of diffusion
diffusion
counterparts
DRAM-
with
projects
their
other
California
system-
TRANUS
consistent
Europe-wide
media
more
still at the core of such
made
ideas
and noise.
In many
of urban
the
MEPLAN,
the
through
to get to grips
of pol-
again
with
America
cities in terms of their physical
processes at work
of local
straints
new ecologi-
of the impacts
models
for
in
by new road build-
set against
and predictions
In
and
Federal mandates
generated
of sprawl
considered
another
use-transport
to assess the impact
EMPAL and MEPLAN frameworks
data
land
raised the need for some large-scale
Traditional
seriously
with
[I9941 efforts
ideas which
[Schock,
to
that takes place. Insofar
ones
municipalities
atic assessment
the growth
for dealing
the problem
cal concerns
models
ly dynamical
on local communities
ing while
urban
concerns
original
Survey
of treating
generally
which
by the late 1980s. the need for sys-
was evident
the US required
for operational
in North
closer
new ways
current
automata
and sprawl
much
growth
is in the spate of urban
US Geological
also for
phology.
with
the
different
modelling
to Chapin’s
mushroomed
of
within
senses, Landis’s
based on cellular
of the
a function
for
strongly
types of infrastructure In the
concern In some
suddenly
developed
1994;
This
the efforts
if any of
Longley,
accessibility
a much closer
models
20001.
being
and
sprawl.
models
kinds of
of efforts
growth
resonates
urban
growth have
city.
time
although
or focus as a brief review these
the
CUF model
system
different
explored,
with
example,
way
of
involving
local
time,
infrastructure
through up
up, and on the way
many
being
parts
and
way
Such is the
and although
in social
processes
on the
through
physical
agents and actors are
the bottom
theory
place
has
This is bound
focus
from
maintain
complexity
individual
global
emerge
to ideas
of urban
Such models
twist
far-from-equilibrium
imply.
disaggregation
states to the point where actions
back
the
such signatures
yet further
being
in
are not the
a further
has moved
scaling
that
outcomes
decade,
Volume 3 - issue 3 - 2001
l
the
capacity
model’s scenarios
two
model
is
con-
dynamics for growth
1990 and 1995 from the set of seed sites which
the development
from
1985 to 1990.
Models planning
In a paper
such as this, it is not possible
any of the models illustrate
actually
developing
software
widespread
exploration
the software the
shift
that
from
occurred
in the last 40 years in
data
which
can
of new
model
forms.
we have developed
simulation
been a sea change
to
in the
towards
which
has
as there
has
representation
it to say that
of cities that we currently
cy while
that
which
ly timely
of the field
review
conclusion
approaches
that
policy
REPRESENTATION:
respect
more
poii-
most represent
a
A particularto operational
by Schock
Agency
based on GIS might
able for urban
of urban
popular.
provided
Protection
are
respect to what
we require
with
been
for the Environmental the
with
is no longer
has recently
to draw
the mathematical
have been explored
way of thinking modelling
but
have at our disposal
cities and what
those
tack
representation
in the millennium,
they are inconsistent
about
through
we will change
in spatial
models
we know
of
representation
suffice
not very good,
for
Moreover,
last 20 years,
digital
this revolution
be used
is characteristic
and GIS. In the next section,
and examine
how
Figures 1 and 2
although
come
and
in this field
graphics
work
far we have
how
to present
[ZOO01
but it is hard not
modest,
data-driven
in many ways be prefer-
analysis.
MODELS
AS DATA
AND
THE GIS
REVOLUTION The problems
that
led to the demise
tion of urban models
produced
ting edge to urban planning inability
to deal with
nature
of computer
changing
computer
technology
development
would
of large-scale
enshrined
rn the history power
were during
the
the
applications
prietary
and ported
els were conspicuous
became
on a chip’
be a problem
models.
Moore’s
of miniaturisation, roughly
predominant
1980s when
models
197Os, it was clear that
198Os, graphics
tion software form
late
(and memory)
and by the
becoming
of urban
never again
computer months,
in the 1960s was already
the ‘microprocessor
and by the
power
cut-
large data systems and the unwieldy
but in 1971,
was developed
digital
in the 1980s and 1990s. Our
as the first generation
operational,
of the first genera-
a very different
many were
now
suggests
that
doubles
and text
every
In fact,
and worksta-
being translated
to microcomputers,
by their absence.
18
processing
applications.
main-frame
in the
Law,
into pro-
urban
Momentum
modin the
field was at such a low ebb that there was little
activity
in
such development
woke
up
and
by the
time
the
field
New simulation diffusion using CA.
models: urban growth
FIGURE 1:
based local
again in the mid 199Os, there were few if any applications that utilised
the then current
making
generic
argued
that
applications
the market
for such development did not attempt
power
of microcomputing
widely
available.
now are such efforts
what
data.
In many
software
in this way,
user interfaces
that became
and
information
and the
available.
digital
Only
their
under way.
planning
257
tools
based
ways,
the
methods
for
was the development
of
on techniques development
systems was an obvious
world data
of computer
in fact happened
but those who were still in the field
they failed to utilise new graphical range of map-based
of the development
modest
software
was too
In terms planners, more
their
models
in be
small
to open
for urban
It might
for the processing
has very wide
of
to the
geographic
step in moving
to a
of maps and storage
applicability,
per se. GIS originated
nearer
from
much
large
a synthesis
of
than
of three
Models planning
JAG
been a central
concern
were
198Os, packages
the early system,
were
being
marketed
planning,
if only
But once desktop
1990s coinciding then
of that
most
field.
of maps as in SYMAP (SYmbol-
based on such associations.
ous uses in urban use mapping.
Volume 3 - issue 3 - 2001
in the development
In fact early line plotting MAPping)
l
By the late
which
GIS began
with
had obvi-
for automated
the Windows
municipalities
began
land
in earnest
in
operating
to employ
such
tools. The functionality
within
kinds of urban
analysis
quite
methods
elaborate
GIS for
most
functions
specialist
plug-ins.
important
in that
process,
particularly
tion,
that
plug-ins
require
those
data
dealing
with
rarely
presented
that
adding
functionality that
controversial
sation permits
to the archiving
technologies abstracted their
areas - from
involved
developments
polygons,
raster
cartography duction
notion
where
and
the concern
so on,
many
that
points,
from
lines,
was the automatic
pro-
and
planning
data
layers
where together
-
data
process,
upon
that
data.
to the building
as the new
Business
District
Bristol can be displayed,
be used
uses at difdesign
level. Much
for use of digital of their
problem
use of that
outcomes kinds
of data
at the fine
types of activity of
the
in a delib-
using GIS, thus
3, we show
smoothed
in and
model
are then
in the digital
In Figure different
the
might
and
In terms of the
to the local urban attitudes,
of
the organi-
of their system as a model
many
the
GIS and range
bases in more
GIS has many
users to think
activity
had
258
Central
with
and
applications
services.
the users to be expert
associated
is a mar-
their
of map data that
to think
Moreover,
of planning
on
modelling esoteric
of tools
planning
process where
upon.
requiring tion
intent
there
across
scale can only be accessed and unlocked
scales,
layers in the landscape
professional
erated
range
even down
requires terms,
problem-solving
computer
of maps across many
architecture
associating
base technologies
in how to represent
ideas in representing
landscape of
data
structures
and generalisation
and from from
spatial
1 and 2 is still
although
where
planners
the regional
of this usage depends
related
the
routine
planning
scale and perhaps
used was
mathematical
as well as professional
scales from
the
In fact, the kinds
in Figures
municipality’s
terms
professional
FIGURE2: Simulating sprawl in the Ann Arbor, MI region
the software
part of GIS are issues involving
of a planning
ferent
for
use.
general for public
plug-ins
in Urban
to develop
offer
from
and
we will
or more ago
GIS vendors,
to note
software
tasks. Clearly
which
they still see as somewhat
in their
It is important related
calcu-
functions,
a decade
for planning
ket, are less inclined applications
of the
Microcomputers
we illustrated in that
and loca-
Most
based on spreadsheets.
available
is
and network
are few
Planning and Management where of software
through
approach
growth
and viewing
in his book
non-graphical,
layer
to 3-dimensions
kinds of tools that were [I9871
on the process
be added
capabilities.
As yet there
from
maps of
areas of the planning
accessibility
extensions
below.
by Brail
the
interpolation
recently
illustrate
to
are many
involve
spatial
numerically
such overlay
to date
lations, more
there
related
Apart
different
need
However
and
limited.
for overlaying
data layers and for operating of overlay,
modelling
is still quite
English
and thence
manipulahow
data
defining town aggregat-
of
JAG
Models planning
ed into
layers that
requires
for smoothing many
are used in defining
considerable
knowledge
that
are embodied
contemporary
professionals In short, various
One of the
reasons
can be adapted are only
why
urban
designers
design
does
have
been
the
hard
at the most local level. Moreover,
GIS [Batty
et a/, 1999b].
ous kinds
of media
use of these data, we need to think
town
displays
into
these new digital bottom-up
tions for a typical
planning
so
implies
ways.
models and
provided
by GIS, which
of visual and statistical
most
appropriate
in planning. provides
applicability
use of
Moreover
a constraint
and
and functions
are left to the user who
also
this which
feasibility
in
these low level tools is engaged
in the
problem. to low
place
fashioned around
of
that
terms.
contexts,
has and
urban
planning.
employed reflected
a toolbox
although
ago,
to
all based
represents
for
represent
of applrcations on
the
the soft-
within
the most widely
mathematical
et al [1993]. based
other
be used extensively
Spreadsheets tool
taken models
is not
GIS now
FIGURE 3: New models of urban data at the fine scale: defining a central business district through diverse data layers which define relevant urban activities in Bristol, UK.
259
have largely
mathematical
or more
continues
In the variety
as those
which
focus of this kind of basic tool,
generic
Klosterman
such
models operatronal
GIS. In fact
ware
by
level
those
a generation
most effective
set of func-
task but in providing
the
technologies
relevance,
in ingenious
as to how to combine
the
but its real power
a purpose-built
models
marks
3 - Issue 3 - 2001
mathematical
that decisions
This shift
into
GIS for
in more abstract
software
to a range
use of software
planmng
and the
and technologies
of development
in provrdrng
probably
to
how vari-
desktop
GIS of course can be used in much more routine is not so much
can be linked
by
they are
of Wolverhampton,
of the problem
such as in the control
these to the data
Volume
data
multimedia
In Figure 4, we show
can be added
English
linking software,
is because the subject
not embrace
too of how easy it IS to embed
of the
limited
to other of building
In turn
Increases
such technologies
of their
that is available
centre
of
requires
this kind of usage.
which which
it also
can be linked
In this, the prospect
of the user.
slow in adopting
unaware
a toolbox
of applications
which
This
operations
in the functionality
GlSs but
are able to exploit
GIS provides kinds
the imagination
matter
desktop
who
boundaries.
of statistical
l
modelllng
collected
Recently
smoothing
together
various
GIS Maplnfo
as
plug-ins
have
been
and aggregating
Models
planning
linked
JAG
to the spreadsheet
ware
continue
and related
tools.
ly elaborate
models this
Many
link
of possible numbers
possible
a spreadsheet of other
applications
reflects
are enormous
urban
planning.
such
as the
building
software
to fashion
opportunities
Other
software
too
development
from
Systems
such as that
contained
different
data
are
is becoming
possible
media.
now
being
are so
over the net and in various is beginning what
important
-
to change
is communicated
penultimate
section
for the develop-
are opening
up the planning
in the domain
tion
of note
graphical
Dynamics-like
we turn full circle and show
sentational into these
the
point
use of icons,
programming
low-level
of view
modelling
it is worth
envi-
of representation.
MOVING
MODELS
TIONS,
Mathematics.
and data are
to collapse
noting
environments
in turn
changing.
contexts
to these
we examine system
is but
- and this
is simulated
is involved
this, we must review
new forms
accessed desktop
and
in model
issues in the
how
to wider where
computers participa-
we stand on
program-
how this new repre-
is beginning
return
where
in and
model-
programming
within
use. We will
models
models
and
what
but also who
design
but before
way
on the
immersive
are in the field.
of
only
not only
IMAGES,
BACK
DIGITAL
The idea that computers Before
not
such extensions,
and
generic
the
communicated
increasingly
3 - Issue 3 - 2001
open to every other
to develop
Moreover
media.
to the
is worthy
of
ming such as STELLA to fully-fledged ronments
there
in comparison
and applications
various
many
such as Excel
and users there
ment of new software
and it is becoming
the fact that the range
is so great
of researchers
every piece of software
extreme-
graphics
and the fact that
probably
Few have the time and ability and there
to develop
within
abound
kinds of soft-
yet more statistical
to a variety
possibilities
few applications
to embrace
It is now
and
to
Excel and these
to expand
Volume
l
that
many
which
pioneers
of the first
of
machines
be ‘programmed’
imply
Increasingly
FIGURE 4: New data models for urban design: articulating
used oscilloscopes
Indeed
REPRESENTA-
be used for graphics
of the computer for monitoring
and realised
to produce
the very first arcade
ICONS:
TOYS
could
back to the very beginning
back
TO
that
pictures
game
‘Pong’
the
the workings
the scopes could of various
kinds.
commercialised
the local design factors digitally in the town centre of Wolverhampton,
260
goes
era when
UK
JAG
Models planning
in the early
1970s by Atari,
ever illustrated television
in the early
revolution
to
raise
become
the
attempts
at putting fly
to
it took
the
digital
level iconic
go back to earliest
through
what
of
and
was possible
kinds of 30 technologies
enables
it has
the
in
first
ways
of the
but widespread
did not occur
of
use of
ingenious
quite
data
began
to
almost
the
market.
and
city
models
and with
navigation.
superficial
However
and their
porary
such design
CAD models to
the
in
various are
offline.
them
urban
geometry.
that
they
form
than
that
associated
rarely
use in anything
building
other
than
in that
all
dominant
urban
design
from
which
is returning
to these
digital
icons is not from
GIS. 3D visualisations
viewing ment
spatial the
extensions into
thrust
which
represent
and in many
2D map view.
30 where
FIGURE
data
- data
enable
Many
GIS packages
2D map
the 3D content
5: Back to lconic modelling:
another
data
to
-
review
population
in the
In Figure
London
3D Analyst
to
a fursuch offer the
and other 5, we
Borough
as a 2D map
is
which
that
CAD models
rather
in digital
than
form
of
and
3D
is part of
applications
various
spatial
database means
being
automated
while
height
such
of
area around
comple-
where
have
sorts.
the Cathedral
the several thousand
attribute.
technology.
important
through in Figure
is also imagery
remote
of the
in the City (of London)
have been taken
This is the state
sens-
6, we show
GIS, a 3D visualisation
points
the
construction.
sourced
itself
of 80
GIS and
representing
photorealistic
St. Paul’s Cathedral blocks
with
model
although
impacts
Upwards
various
For example,
using desktop
the building
LIDAR data,
be extruded
through
to accessing
aesthetic
designs.
technologies for
represented
Users now require
exist worldwide
data is being
ing of various again
no longer
as a means
to their
now
of 30 digital
[Batty et a/, 20011
for solely generating
visual form
serious
once
review
of London
Realism in terms of rendering
and
way
a worldwide
edge of such technologies.
predominant
CAD but
now
is some spatial
For
taken
users to explore
scenes.
is illustrated
we undertook
models
data,
and
planning
senses these
density
for the Corporatron
or providing
limited. The
data.
correlations
different
using the plug-in
iconic
struc-
rendering, design
enables
GIS ArcView.
the cutting
for visualising
have any content
that
block model
and concluded
the idea
The geometric
with
In the is often
then some packages
and examining
which
is not possi-
are very effective
into the scene. If several
together,
between
and
and if this is used, then
the desktop
models
to presen-
limited
patterns
population
Recently
relatively
In fact, contem-
is use them
in a way that
of any activity
hotlinking
pan
inter-
used for display
models
of cities are rather
tures
their
3D
such CAD invokes
is always
other hence
such
many
zoom
at multiple
can be coded
Westminster
developed
usage has been confined
one can do with
changes
been
of the web,
Although
computer
For example, have
the growth
purposes.
of design,
personal
such as VRML are being
net equivalents
tational
as the
invade
architectural AutoCAD
as soon
easily
to
30 views
by brushing
3 - Issue 3 - 2001
can be more
ability
In fact,
density
ther variable
show
that
the
associations
began
and
looking
Volume
in data
the
to be examined
scenes are displayed
recently. CAD
errors
while
mark the thrrd dimension
early
until
and
30
detail
example,
downtown in the
in
ble in a map view.
days of CAD. The
Merrill
patterns
visualised
the micro-
model
buildings
Owings
Spatial
on US
where
technology,
the traditional
by Skidmore,
1980s showed these
first shown
1950s. Although
dominant
frame
Chicago
computer
graphics
city into a computer wire
was one of the first graphics
from a digital
l
being
generated
from
crude
constructed
from
by the laser-based
of the art and very shortly
density in Westminster (London) generated as block model within desktop GIS
261
Models planning
JAG
such models will be generated on the fly as satellite data at the requisite level of resolution becomes available on a daily basis.
Volume 3 - issue 3
- ZOO?
the level of representation and it implies that the kinds of models that are suitable are those which are less abstract, more micro in scale and form, as well as dealing with routine dynamics processes that might be captured dynamic reaiisations of the 2D map or 3D geometry
One of the main themes in this paper involves the way the idea of the computer model of the city has changed during the last 50 years, and our implicit argument that
in in
question. For example, work at the Environmental Simulation Center (http://www.simcenter.org/ Projects/ CommunityViz/communityviz.html~ in New York City for the town of Scutney in New England has developed a
we have come full circle is clearly being borne out by the flurry of 30 digital block models of cities that are currently being constructed. Yet this circularity is never what it seems. This is no mere return to a world of the archi-
visualisation of an agent-based model of community development, tying its mathematics back to data as well as to 3D icons. We have ourselves designed various models of the movement of agents in shopping centres and
tect’s models where the focus is purely on visual evaluation. As we shall see in our concluding section, access to these new digital icons is very different from the precomputer age. The visual simplicity of their form belies their true purpose which is much more likely to be as an
in galleries and have visualised such movements in 3D. In our models of movement in the Tate Gallery, we have visualised agents moving in both 2D and 30, examining different digital visualisations and simulations of the same system [Batty et al, 19981. We show some of these in Figure 7 but our point is that the argument of this
interface to a complex and interrelated data base where 3D visualisation is but one way of accessing the data and managing this complexity.
paper is by no means complete. In the next decade we are likely to see many new fusions and extensions of the mathematical, representational and data models which we have through different kinds of digital media
l-lowever there is another set of applications which is likely to convolute these examples even further. We have already implied that new functionality is being introduced into GIS which is extending their use from representation to simulation. The same is occurring in 3D. Already there are traffic models where the representation enables actual car movements to be simulated mathematically and visualised in 3D while there are experiments to show how various kinds of model process can be embedded within 3D virtual environments of the kind visualised above. This is the point at which the models themselves must match
FIGURE6:
l
MOVING MODELS BACK TO USERS: PARTICIPATING IN DIGITAL DESIGN In public policy particularly, there has always been concern as to the relevance and remoteness of the science with respect to the kinds of advice that such science provides for issues that affect a wider public. However
A GIS 30 block model of St Paul’s district in the city of London with height data from LIDARdata
262
JAG
Models planning
realrty
in a digital
any problem
different There
world
can be quickly
computer
models
interpretations
of data
models,
individually
As we have noted, being
ported
to the web.
software Internet
ed in thus paper, the
regeneration Hackney. which
The Hackney
is built
around
environmental designs imtiative
aimed
images
from
of
is a project
approaches
to raising
non digital
maps and
by a complementary
participation
the process. this
of
Borough
in the local community
supported
and icons to support some
city context
Exploratory
through
at public
by a
is possi-
problems
in the London
products
being
about
conventional
awareness
and related
this is now
upon
of what
used In an inner
Building
project
is
that we have present-
of awareness
and deprivation
kinds
visualizations
and acted
As an example
are being
of others.
of varrous
GIS and CAD is being
can be viewed
development
own
burld therr own
and this IS delivering
of interests.
enor-
their
or in the presence
ble, many of the visual techniques for
of users
web provides
and rn prrnciple,
established
range
wide
a range
users to generate
desktop
and map data that wide
on
from the use of
in which
but, the world for different
gradually
perspectrves
fashioned
forums
mous potential either
different
based tools.
are as yet few
can build
where
Volume 3 - Issue 3 - 2001
l
and digital
using digital
In Figure
which
CAD and GIS can be used to unlock
data
8, we show
reveals
how
VR,
data and to deliver
this to users across the web. This data vanes in scale but the intent
is to educate
and communicate
cern in the local community. to provrde way they
users with that
the interests
building
capability
are being
built
as yet
but in the
to put data together,
models
and beliefs
by local planning FIGURE
We have not sought
model
are encouraged
an implication
issues of con-
there
which
is
reflect
of those who are being affected
issues.
7:
Spatial modeiling at the finest scale: simulating human movement in 25 with visualisation in 30. In Figure could attempts
at developing
systems
In the
wider
attempts
at
dynamics
models
context
providing
sion-makers
al interfaces
might
approach
However,
those to
the
with
degree
uncertainty
of
this
more than one model ‘counter [I9761
perspective modellrng’ more
close
The notion
associated
anced
jointly
but in general,
than
same
with
lated
to the
science
are able
problem
IS usually
IS also
of participation. systems
required
policy-making. proposed
change
The high
terms of what
to provide The much
cy whrch
a bal-
process more
the earliest
that
texts.
of
The only we might
saw their
ourselves
limits
In the next 20 years, there
of ideas which
of a
planning’
will
broaden
even further.
of
up a variety
of
and modified There
from
is a
the digr-
and all promise
to
In public affairs and are in
and this IS a far cry from
in planning
use in purely
simu-
kinds
to such possibilities
imagine
use of computers
et a/
263
constructed
the way we involve policy.
emerges
- the
with
not been possible. which
IS being
be tested
their
fiction
- but these
and open
might
exten-
users might
but real individuals
possrble
movement
alongside
like science
interactions
simulations
whrch
of
in which
environments
have hitherto
urban
is such
such
of possibilrtres
intimately
by Greenberger
20 years ago is now
tal world
to
more than
models
and an obvious
ways
This sounds
are now
in ways that cornucopia
inter-
involves
and their
ways in which
how worlds
of remote
agents
environment
using visu-
in which
enter selves.
Juxtapositions
deci-
and the new con-
too of providing
in human
first
simulated
such efforts
modelllng
a context
process
on
themselves
Some systems
the new focus on vrsuali-
is at last providing than
limited.
illustrated to virtual
sion of such worlds
of urban
in which
be developed
up data and symbolic
make an impact. one
been
GIS and VR are providing
cern to open ests other
have
have been tried
which
models
environments
have been laboured. sation
mathematical
7, we
be linked
passive, promises
the concept
and public
poli-
top-down
con-
to be a fusion of ‘models
In
JAG
Models planning
FIGURE8: Bringing digital data and models to users: participation interactive
CONCLUSIONS: PLANNING
levelled to
get
AND
MODELLING
of models
is no more
that
trenchant
we have developed that
those
at every area of the human to
fraught
FOR
CITIES IN THE 21ST CENTURY
The critique paper
MODELS
in environmental
grips
with
social
that
have
sciences.
complexity
phenomena,
in this
in terms
as urban
that supposes
that the human
condition
can be bettered
by interfering
in the
of urban
development,
process
and
the
shown.
how
cities function
researchers
innovate
264
We still are
complexity, wisdom systems and
do
not
Issue 3
more
models
as those
2001
good
there aware
theory
are signs of
their
have
been
which
devel-
form
the conven-
economics.
Moreover,
as their
populations
are ever changing new
last 50
that
of contemporary develop
of the
have
and although
becoming the
are as arbitrary
human is
3
as the experience
have
tional
ideology
difficulty
about
oped
of its
manifestation
with
intrinsic
Volume
issues through the Hackney building exploratory-
years that
been
The effort
l
sets
of
preferences
and
Models
JAG
ptanning
tastes,
thus
applicable.
laying
waste
The current
sense of complexity of the fact that ingful that
concern
it is virtually
that
once
for theories
and change
predictions would
to theory
seemed
that
environment
make
potential
to make
mean-
urban
form
the basis of medium
in using
models
that
or long term
have
material
we have addressed
in planning,
circle and it worth
pausing
learned
to the real thing,
to briefly
along
the
digital
iconography.
To progress
cities involved
reflect
way.
forms
of their
with
come
full
on what
we
based
on
Models
by models
this far, the first applications
of space together, across
visual
simulations
but now
the
imagery
realm
to
is now
a cliche
comparisons
the
ing of how has been
much
a cautiousness
knowledge
both
little
that
doubt
our
that
new
these will
than In fact,
new plane
in terms
ops theory,
more
in the physical
end,
many
urgently that
with
data
and devel-
to the way
sciences
is required
large-scale
are being
techniques methods
that
ideas
they
and processes.
presented
to
and the data
A
D., 1992. The Dynamics of Crtres. Routledge,
London.
Echenique, M H., 1972. Models. a drscussion. In: L Martrn & L March (Eds), Urban Space and Structures. CambrIdge University Press, Cambridge, pp. 164-174.
to
and
Echenrque, M.H , 1994. Urban and regional studies at the Martin Centre Its orrgms, its present, Its future. Environment and Plannmg 6 21: 517-534.
rich
Forrester,
scales,
Counter-modelling
are now essential
Dendrinos,
and in hierar-
different
of Form
De La Barra, T., 1989. Integrated Land Use and Transport Modellrng. Cambndge Umversrty Press, Cambrrdge.
building, seeks
a
Chapm, F.5. & 5 f Weiss, 1968 A probablllstic model for resrdential growth. Transportation Research 2. 375390.
the basis
As ever, the
which
Systems.
Brewer, G.D., 1973. Polrtrcrans, Bureaucrats and the Consultant’ Critique of Urban Problem-Solving. Basic Books, New York.
for searching
of theory
Sustainability
Brawl, R.K , 1987. Microcomputers In Urban Planning and Management. Rutgers Universrty Press, New Brunswick, NJ.
seek to explain.
in parallel
tailored
sys-
To this
in any case are
in planning
and models
in parallel
which
by the ESRC NEXand
Ben Akrva, M. & S. Lerman, 1985. Drscrete Choice Analysis: Theory and Application to Travel Demand MIT Press, Cambndge, MA
in unlocking
provide
and association.
to a
data
handled.
are essential
is a strategy
prediction
sectors
been
it treats
on the very world
modelling
have
of the way
of causality
problems,
but
move
wait for new theories
different
we
financed
Complexity
Batty, M , D Chapman, S. Evans, M. Haklay, S Kueppers, N. Shiode, A. Smith & P Torrens, 2001. Visualrsrng the city: communicating urban design to planners and decisron-makers. In. R. Bra11& R. Klosterman (Eds), Planning Support Systems. Integrating Geographic InformatIon Systems, Models and Visualrsation Tools. Rutgers University Press, New Brunswick, NJ/ ESRI Press, Redlands, CA, pp. 405-443.
intensive
must
in the data that might
techniques
than
required
for
Batty, M., M. Dodge, B Jiang & A Smith, 1999b. GeographIcal information systems and urban design In: J. Stillwell, S. Geertman & 5. Openshaw (Eds), Geographical Information and Planning. Spnnger Verlag, Herdelberg, pp. 43-65
is
studies
data sets and developing
and
There
in much more
anything
part of this paper
for new notions
chy,
are
Network
Batty, M., Y Xie, & 2. Sun, 1999a. Modelling urban dynamics through GIS-based cellular automata. Computers, Envrronments and Urban Systems 23: 205-233.
marks the challenge.
of the representational
for new patterns
modelling
our
urban
akin
tems
Thus what
in using
even more inadequate
of cities
environments
dependent
has
and prescriptively.
and it is this that
theories
cannot
is
this
the tools we now have are all the better
used to hitherto.
complex
us pause
need to be developed
in the latter
of cities
with
has been partly
SUS Project:
Batty, M., R. Conroy, B. Hrllrer, B Jiang, J. Desyllas, C. Mottram, A Penn, A. Smith & A. Turner, 1998. The vrrtual Tate. Workmg Paper 5. Centre for Advanced Spatial Analysis, Universrty College London, London. (http://www.casa.ucl.ac.uk/tate.pdf)
our understand-
and
This research
Batty, M & P. Longley, 1994 Fractal Citres, A Geometry and Function Academrc Press, San Diego, CA
are
this era. Complexity
makes
for this but they now appear they once were
in
of information,
today
predictively
this
to one
the planning
during
thinking
up which
of energy
in terms
enriched
come
arrange
of
and policy-making.
Batty, M., 1994. A chronicle of scientific plannrng, the AngloAmencan modelling experrence. Journal of the American Planning Assocratron 60: 7-16.
It is almost
sums
but it is very clear that
dominates
world
understand-
generation
Batty, M., 1979. Progress, success and farlure in urban modelling. Envrronment and Planning A 11, 863-878.
from
processes
a world
in terms
we need to approach
what
data
next
Batty, M , 1976. Urban Modelllng: Algorithms~ Calrbratrons, PredIctIons. Cambridge Unjversity Press, Cambridge
exists of
realism
which
between
more explicable
not very fruitful
prospect
relative
cities function.
past was best explained
Thus,
the
Batty, M , 1972. Recent developments In land use modellrng, review of bntish research Urban Studres 9: 151-177.
to
where
to social and economic
provide
absolute
the
of
lie at the basis of how
impossible
that
mark
Allen, P.M., 1997. Crtres and Regrons as Self-Organising Models of Complexity. Gordon & Breach, Amsterdam.
the same kinds of visual
much more abstract
synthesising
the
basis for this new
REFERENCES
based
focus was upon the map and the processes that held this
change;
will
has all the
(L326253048).
visual superficiality
have been supplanted
on ethereal,
which
the
here
poli-
here is that
we have almost
icons, close in terms
obvious
research
have sketched
3 - Issue 3 - 2001
ACKNOWLEDGEMENTS
One of the themes
kind
to provide
Volume
for such systems or at least predictions
cy-making.
might
we
ing. This we suggest
on all scales is indicative
impossible
which
l
265
J W, 1969. Urban Dynamrcs. MIT Press, Cambridge, MA.
Models
planning
JAG
l
Volume
3 - Issue 3 - 2001
Greenberger, M., M.A. Crenson & B.L. Cnssey, 1976. Models in the policy process: Public decisron-making In the computer era. Russell Sage foundation, New York, NY.
Wilson, A.G., 1981. Catastrophe Theory and Brfurcatron: Application to Urban and Regional Systems University of California Press, Berkeley, CA.
Harris, B., 1968. Quantitative models of urban development: their role in metropolitan policy-makrng. In: H.S. Perloff & L. Wingo (Eds), issues in Urban Economics. The Johns Hopkrns University Press, Baltimore, MD, pp. 222-222
Wilson, A.G., 1998. Land use-transport models: past and future. Journal of Transport Economics and Policy 32. 3-26.
Harris, B., 1994. The real issues concerning Lee’s ‘Requiem’. of the American Plannrng Association 60: 3 l-34.
Wilson, A-G., J.D. Coelho, SM. Macgill & H.C.W.L. Williams, 1981. Optimisation in LocatIonal and Transport Analysis. John Wiley & Sons, Chichester.
Journal
Wlttgensteln, L / 1921 (revised edition 1961). Tractatus Philosophicus. Routledge & Kegan Paul, London.
Hill, D.M., 1965. A growth allocation model for the Boston region. Journal of the American Institute of Planners 31: 11 I-120. Klosterman, R.E., 1994. Large-scale urban models: retrospect and prospect. Journal of the Amerrcan Planning Association 60: 3-6.
RESUME Le terme “modele” est entre dans le dictionnaire au tours des annees 60 quand soudainement I’idee de representation symbolique de systemes complexes a vu le jour. Cecr etait en grande partie due aux ordinateurs qui etaient capables de traiter automatiquement de grandes series de don&es ainsi qu’a un changement fondamental de notre comprehension des systemes dans le domaine de la science ou de la societe. Comme les ordinateurs ont envahi tous les domaines de notre monde, I’idee du “mod@le” n’avait plus le meme pouvoir attractif. Les modeles sont partout. Cependant pour des systemes urbains, le role des modeles a enormement change. Avant 1950, des modeles urbains designaient des representations architecturales de leurs formes physiques. Des modeles mathematiques ont domine les annees 60 et 70, puis il y a eu Ies modeles de donnees qui ont domine au tours des an&es 80 et 90. Mais maintenant I’accent est de nouveau sur une representation tradrtionnelle des villes par des modeles de forme tndimensionnelle, annoncant ainsi un retour des modeles d’icones qui ont domine la planification physique jusqu’ici. Dans cet article, nous passons en revue le developpement de la mod~lisation dans la recherche urbaine au tours de la moitie du dernier siecle et nous suggerons que des nouveaux types de synthese sont maintenant possibles.
Klosterman, R E., R.K. Brawl & E.G. Bossard (Eds), 1993. Spreadsheet Models for Urban and Regional Analysis. Rutgers University Press, New Brunswtck, NJ. Landis, I.D., 1994. The California urban futures model: a new generation of metropolitan slmuiation models. Environment and Planning B 21: 399-420. Lee, D.B., 1973. Requiem for large-scale models. Amerccan Institute of Planners 39: 163-l 78.
Journal
of the
Lee, D.B., 1994. Retrospective on large-scale urban models. Journal of the American Planning Association 60: 35-40. Lowry, I.S., 1964. A Model of Metropolls. Corporatron, Santa Monica, CA.
RM-4035-RC.
The Rand
Lowry, I.S., 1965. A short course in model design. Journal of the American Institute of Planners 31: 158-I 66. Portugalr, J., 2000. Self-Organisation Heidelberg.
and the City. Springer-Verlag,
Putman, S.H., 1983. Integrated Urban Models: Transportation and Land Use. Pion, London
Policy Analysis
of
Putman, S-H., 1991. integrated Urban Models 2: New Research and Applications of Optimisation and Dynamics. Pion, London. Schock, S., (Ed.), 2000. Projecting land use change: a summary of models for assessing the effects of growth and change on land use patterns. EPA/600/R-00/098 National Exposure Research Laboratory, Office of Research and Development, US Environmental Protection Agency, Washington, DC.
RESUMEN El termino “modelo” entro en el Iexico en 10s arias 1960, cuando se desarrollo repentinamente la idea de representar simbolrcamente sistemas complejos. Esto se debe tanto al hecho que las computadoras habian alcanzado el punto de manipular amplias series de datos en forma rutinaria coma a cambios fundamentales en nuestro entendimiento de 10s sistemas compleJos en la ciencia y en la sociedad. Como las computadoras han penetrado cada esquina de nuestro mundo, la idea de modelo ya no tiene mas el poder de atraccion que tuvo en otro tiempo. Modelos se encuentran en todas partes. Sin embargo, en el case de 10s sistemas urbanos, el papel de 10s modelos ha cambiado substancialmente. Antes de 1950, 10s modelos de ciudades se referian a representaciones arquitectonrcas de su forma fisica. En 10s arias 1960 y 1970 dominaban 10s modelos matematicos, y en 10s arias 1980 y 1990 comenzaron a dominar 10s modelos de datos. Pero ahora, el enfoque volvio a las representaciones tradicionales de las cuidades coma modelos drgitales tri-dimensionales, anunciando de esta manera un retorno a /OS modelos figuratlvos que habian dominado la planificacl~n fisica hasta ahora. En este articulo, se analiza el desarrollo de la modelizacron en investrgacion urbana durante el uitimo medio siglo y se sugiere que nuevas formas de sintesis son ahora posibles.
Simon, H.A., 1957. Models of Man. John Wiley & Sons, New York, NY. Srmon, H.A., 1977. Models of Discovery and Other Topics in the Methods of Science. D Reidel, Dordrecht. Simon, H.A., 1982. Models of Bounded Press, Cambridge, MA.
Rationality
(2 volumes) MIT
Simon, H.A., 1989. Models of Thought, Volumes 1 and 2. Yale University Press, New Haven, CT. [AUTHOR: I have just given one date although the first volume was published in 1979, the second In 1989, but I think It IS overkill to put them In twice as separate entries] Simon, H.A., 1991. Models of My Life. Basic Books, New York, NY. Waddell, P., 2000. A behavloural simulation model for metropolitan policy analysis and planning: residentral and housIng components of UrbanSlm. Environment and Planning B 27: 247-263. Wegener, M., 1994. Operatronal urban models: state of the art. Journal of the Amencan Planning Assocratron 60. 17-30. Wegener, M., 2001. New spatial planning models. Unpublished paper, available from lnstitut fur Raumplanung, Universitat Dortmund, Dortmund, Germany Wilson, A.G., 1974. Urban and Regional Models in Geography Planning. John Wiley & Sons, Chichester.
Logico-
and
266
