A pluralistic approach to defining and measuring urban sprawl - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 158

development patterns. Examples for each of the three definitions
are provided below.

12.2.2 Definitions based on
consequences of sprawl; sprawl
is as sprawl does

12.2.1 Definitions describing
an urban spatial development
phenomenon

''Ultimately,'' write Ewing and colleagues, ''sprawl must be judged
by its consequences'' (Ewing, Pendall and Chen, 2002). Thus, the
definition of sprawl becomes the socio-economic or ecological
consequences of a particular kind of urban spatial development.
Consequences might include (1) lack of accessibility between
regions in the urban area (Ewing, Pendall and Chen, 2002);
(2) high rates of driving and vehicle ownership (Burchell et al .,
1998; Ewing, Pendall and Chen, 2002); (3) increased air pollu-
tion (Ewing, Pendall and Chen, 2002); (4) undesirable ecological
impacts, such as impact on ecosystem cycles or species composi-
tion (Perry and Dmi'el, 1995; Cam et al ., 2000; Kreuter et al ., 2001;
McKinney, 2002; Hasse and Lathrop, 2003b; Robinson, Newell
and Marzluff, 2005); (5) consumption of exurban open space
and agricultural land (Burchell et al ., 1998; Hasse and Lathrop,
2003b; Frenkel, 2004; Czamanski et al ., 2008; Koomen, Dekkers
and van Dijk, 2008; Thompson and Prokopy, 2009), and/or (6)
catalyzing socio-economic and racial segregation (Squires and
Kubrin, 2005). Some of these variables can be measured directly,
particularly loss and fragmentation of open space, while others
depend on proxy measures and non-remotely sensed data.
Most, if not all, of these claims are contested. For instance,
Glaeser and Kahn (2004) note that while sprawl and associated
increases in private automobile use may have increased air pollu-
tion, technological improvements in fuel efficiency and emissions
control have led to an overall reduction in most air pollutant
emissions in the United States. The claim of sprawl leading to
socio-economic and racial segregation is also challenged (Glaeser
and Kahn, 2004; Wheeler, 2008). Further research attests to the
potential benefits of sprawl in terms of maximizing consumer
preference, efficient distribution of business and residential areas,
low cost relative to high-rise or high concentration settlement
(Gordon and Richardson, 1997), and increasing species and
ecological habitat diversity (Czamanski et al ., 2008).
The great interest that urban planners, policy makers, scholars
and activists share regarding sprawl is, to a large degree, derived
from opinions regarding how a city should develop spatially, and
whattherole(ifany)theplannerandpolicymakershouldserve
in promoting or preventing sprawl. Researchers have noted that
the debate around sprawl is often the result of its ideological
framing (Burchell et al ., 1998; Chin, 2002; Hasse, 2004, 2007).
Thus, some researchers and activists define sprawl in a pejorative
way in order to advocate or oppose a particular policy or plan.
One's description of sprawl characteristics can thus be seen as
a subjective extension of values-laden planning goals; that is,
sprawl is in the eye of the beholder. Opponents of sprawl define it
in terms of what it is not: highly centralized, compact cities with
mixed land uses, whose transportation systems de-emphasize
the role of the private automobile in lieu of public and/or non-
motorized transportation (e.g., Duany, Plater-Zyberk and Speck,
2000). Advocates of more laissez faire policy approach define
it in a more positive light: benign at worst and the desired
expression of people's residential preferences at best (Gordon
and Richardson, 1997). Simultaneously, these latter scholars
provide research results that challenge the claims of the former
group.

Sprawl is most often considered a particular spatial pattern of
urban development characterized by low density residential and
commercial development. Low density could be considered in
terms of building density or population density. This develop-
ment may be adjacent to existing development, as with suburbs,
or scattered and discontinuous development physically separated
from the central city, as with leapfrog development (Harvey and
Clark, 1965; Downs, 1994; Ewing, 1997; Burchell et al ., 1998; Hess
et al ., 2001; Chin, 2002; Ewing, Pendall and Chen, 2002; Glaeser
and Kahn, 2004; Tsai, 2005; Torrens, 2008). As Chin points out,
both forms of development are classified as sprawl, although ''the
forms and resulting impact are vastly different'' (Chin, 2002).
This is at least partly understood when considering that sprawl
can have different definitions at different spatial scales (Tsai,
2005) - for instance at the scale of a single urban settlement or at
the scale of a region of multiple settlements (see below). Sprawl
is also defined as developed land highly segregated into single
uses (Ewing, 1997). The presence of large blocks of exclusively
residential land or commercial strip development, for example,
is considered sprawl (Chin, 2002).
While there are multiple, measurable characteristics of urban
sprawl (about which we expand upon in a following section),
we note that there are no settled values, or quantitative thresh-
olds, that define sprawl in absolute terms. Proposed absolute
values or thresholds that separate 'good' spatial development
from 'bad' are subject to debate, as is the question of what
residential density constitutes sprawl (Chin, 2002). Superla-
tives are common throughout the sprawl literature, describing
the phenomenon as ''excessive'' (Bruekner, 2000), ''wasteful''
(Torrens and Alberti, 2000) and ''inefficient'' (Fulton et al .,
2001; Peiser, 2001; Frenkel and Ashkenazi, 2008a; Thomp-
son and Prokopy, 2009). Others describe the kind of urban
growth considered to be sprawl as ''dysfunctional'' (Ewing, Pen-
dall and Chen, 2002). But sprawl is clearly a relative, rather
than absolute, phenomenon (Frenkel and Ashkenazi, 2008b;
Bhatta, Saraswati and Bandyopadhyay, 2010). This is explic-
itly recognized in the work of Sutton (2003), for example, for
whom sprawl is relative to an average relationship between
population size and developed area across US metropolitan
regions.
We suggest that one way of working towards consensus on
the matter is to define sprawl as a directional process (Harvey
and Clark, 1965; Hess et al ., 2001), rather than an absolute
state of being. Accepting this, the dynamic temporal and spatial
patterns of urban spatial growth become crucial to measure
and monitor. Noting how these patterns change over time and
space change the debate from one about sprawl (a state) into
one about sprawling (a process). In other words, while we may
not be able to agree that a given density constitutes sprawl,
we can call a process of declining density, for example, as
sprawling.