Rethinking progress in urban analysis and modeling: models, metaphors, and meaning - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 356

TABLE 26.2 Four urban modeling traditions: root metaphors and operational models.

Urban analysis and Related pepper Dominant metaphors Operational models/

modeling traditions world hypothesis measurements

Spatial morphology Formism Cities as Fractals Fractals

Lacunarity
Spatial metrics

Social physics Mechanism Cities as Machines The Lowry Model

CUFM, ITLUP,
MEPLAN,RURBAN,
URBANSIM

Social biology Organism Cities as Organisms Cellular Automata,

Agent-based Models
(ABM), Life Cycle
Analysis (LCA)

Spatial events Contextualism Cities as Arenas Field-based Time

Geography;
Sequence Alignment;
Urban Social Tapestry
Analysis

years can better be grouped into four major traditions - spatial
morphology, social physics, social biology, and spatial events.
Embedded in each of these traditions is a different driving
metaphor urban modelers live by (Table 26.2), corresponding to
one of Pepper's four root metaphors in his world hypotheses.
Viewed from a metaphorical angle, the first generation of
urban models was primarily motivated by conceptualizing cities
as machines following the social physics tradition, whereas the
second generation of urban models conceives cities as organisms
following the social biology tradition. In addition to these two
dominant types of models, urban modelers have never ceased
their efforts to search for innovative ways to better describe
complex urban forms. In addition, the recent emergence of
geographical information systems (GIS) and science (GIScience),
Web 2.0, and user-generated content have also promoted a
nascent paradigm of studying cities as events. As shown in
Table 26.2, Pepper's four world hypotheses can provide us
with a more holistic perspective on diverse urban analysis and
modeling efforts (Wilson, 1967; Batty, 1976; Harris, 1985; Batty
and Longley, 1994; Benenson and Torrens, 2004; Hudson-Smith
et al ., 2009).

theory for residential location all entail certain geometric forms.
In urban geography, Burgess's (1925) concentric rings, Hoyt's
(1939) sectoral radiation, and Harris and Ullman's (1945) mult-
inuclei, and Garreau's (1992) edge city are widely accepted as
standard descriptions of urban spatial structure based on geo-
metric forms, which in turn have been transplanted to social area
analysis from a factorial ecology perspective. Recent debates on
new urbanism and whether urban development should follow
more compact or dispersed/sprawling patterns are also, perhaps
indirectly, a reflection of our obsession for the pursuit of formism
(the ''ideal'' form) in understanding cities.
Although non-Euclidean geometry has been occasionally
mentioned in the literature (Muller, 1982) to depict functional
areas in cities, until recently descriptions of urban forms have
been predominantly couched in Euclidean geometry. Exceptions
like Wilson (1981) attempted to apply Thom's (1975) catastro-
phe/bifurcation theory to study cities generated few followers.
One of the major breakthroughs in the spatial morphology
tradition is perhaps the conceptualization of cities as fractals
according to Benoit Mandelbrot's fractal geometry. A fractal is
a geometric form that has the property of self-similarity - some
part of the form has the same properties of the form as a whole.
In urban studies, cities possess fractal properties in a statistical
sense, meaning that a part of the form has the same statistical
properties as the whole. Batty and Longley (1994) argued that
much of our pre-existing urban theory is a special case of the
fractal city, and they see fractal as the best hope for a holistic
understanding for cities by linking the micro with macro, local
with the global, intra- to inter-urban levels of analysis. To a larger
extent, recent advances in our understanding of the size, scale,
and shape of cities are all closely related to conceiving cities as
fractals (Batty, 2008; Rozenfeld et al ., 2008).
By conceptualizing cities as fractals, we can now better under-
stand the complexity of urban forms in ways that are impossible
using the Euclidean geometry, such as multi-scale measurements
and dimensions between 1 and 2. The fractal dimension is quickly
becoming a major component of spatial metrics for describing
landscape patterns and urban forms (Sudhira, Ramachandra and
Jagadish, 2004; Herold, Couclelis and Clarke, 2005). We can

26.3.1 Cities as forms - the spatial
morphology tradition

The study of urban forms has a long interdisciplinary history
linked to geography, planning, architecture, economics, and
sociology (Morris, 1979; Vance, 1990; Kostof, 1991). This tradi-
tion focuses on the description, analysis and modeling of existing
or ideal urban forms. It corresponds to Pepper's formism world
hypothesis, with a root metaphor focusing on the various mosaics
of geometric forms.
The spatial morphology tradition is perhaps the oldest among
the four. Good city form is often linked to the stars in ancient
China and India (Lynch, 1981). In classic locational theories, von
Th unen's concentric rings for agriculture, Weber's triangle for
manufacturing, Christaller's hexagons for services, and Alonso's