Environmental Engineering Reference
In-Depth Information
Many of the variables used to measure homogeneity/
heterogeneity are again drawn from the discipline of landscape
ecology (McGarigal et al ., 2002). Variables such as leapfrog and
connectivity indices describe the mix of urban ''patches'' within a
matrix of open space and measure the proximity of similar patch
types from one another (Galster et al ., 2001). These measures
quantify the level of scatter and fragmentation of the urban
landscape. When built areas are separated from one another by
open space then the landscape is considered fragmented, which
is another sprawl characteristic (Torrens and Alberti, 2000).
One of the most intuitive measures derived from landscape
metrics is the number of patches of a certain land use type. The
larger this number, the more heterogeneous mix of land use
patches at the landscape scale. Mean patch size takes the average
size of all of the patches of a given land use, and the smaller
the average patch size, the more heterogeneous or fragmented
the landscape and thus the more sprawled (Torrens and Alberti,
2000; Herold and Menz, 2001).
Three additional sample sprawl measures derived from land-
scape metrics that measure patch composition are contagion,
connectance, and proximity. Contagion is the tendency of patch
types to be aggregated (McGarigal et al ., 2002); high contagion
value at the municipal scale could suggest large tracts of homo-
geneous land use or sprawl. However, at the regional scale high
contagion value might suggest a low amount of fragmentation of
the landscape, with built patches clustered and not fragmenting
the open space 'matrix'. Connectance and proximity both com-
pute the functional closeness of patches of similar type, and their
values are interpreted such that greater dispersal of patches (i.e.
built patches in an open space matrix) represents greater sprawl.
Finally, several variables measure the degree of irregularity of
the patch including circularity and edge to area ratio. In terms
of sprawl, irregularity is considered sprawling, with a perfectly
circular patch synonymous to compact development, as opposed
to linear or irregular development (Gibbs, 1961).
of possible trip destinations in a given area; gravity indices based
on the classical gravitation model used in urban planning - the
movements of goods, people and information between different
spatial locations, often referred to as origins and destinations,
and; utility function index gathered from discrete choice models
customary to transportation planning discipline (Torrens and
Alberti, 2000). Degree of dependency on private automobiles
for transport is also considered to be a proxy for sprawl. Where
accessibility is lower, there is a higher reliance on private auto-
mobiles to connect between the residential and other land uses
(Ewing, 1997, 1994; Ewing, Pendall and Chen, 2002).
12.4.2.5 Aestheticmeasures
Sprawl is often considered a boring, homogeneous form of
development (Fulton, 1996; Gordon and Richardson, 1997).
Being subjective by definition, it is difficult to measure and
quantify the aesthetics of sprawl unless by consumer preference
or survey data. Several recent studies have attempted to define
archetypes of urban development or sprawl, such as residential
sprawl or strip-mall sprawl, and to compare various landscapes
to those archetypes. It seems that much work is still needed in
this area (Torrens and Alberti, 2000), and, as noted above, these
measures are less relevant to remote sensing experts.
12.4.3 Choosing among the
sprawl measures
The advantages and disadvantages of each of these sprawl mea-
sures can be considered on the basis of the five criteria outlined at
the beginning of this section. We selected a representative sample
from the span of possibilities and ranked them according to how
well they comply with the five criteria. Our ranking is based on
a comprehensive literature review (and thus the experiences of
other researchers), as well as our own experiences measuring
sprawl and conveying concepts and empirical findings to col-
leagues, students, professionals and stakeholders. Rankings are
on a scale of one to three, with three being the highest ranking
variables for the given criteria. Applicability for different spatial
scales receives its own ranking system; from A to C, with A being
applicable to multiple spatial scales and C to only one scale.
Results are summarized in Table 12.1.
We offer three caveats to our ranking of the sprawl variables.
First, a measurement that receives high marks across all categories
does not necessarily make it the best measurement for all case
studies or for measuring all aspects of sprawl. The sprawl quotient,
for example, has disadvantages (noted below) that are only picked
up in one of the five criteria and ranks highly in the other four.
Further, there are measures that received low marks, but still
provide important information regarding sprawl - sometimes
only at particular spatial scales or for particular places, but useful
nonetheless.
Second, while the user can choose from among these variables
in a way that suits their specific research question, data sources
may also determine the variables selected. The relevant consider-
ations here are (1) availability of historical data and (2) the scale
of resolution. With regard to the first consideration, as we have
emphasized, sprawl is both a state and a process. As a process,
directionality of trends is important. Therefore, it is crucial that
12.4.2.4 Accessibility between
residential, commercial and
business areas
Sprawl is defined as a condition of poor accessibility, followed
by the massive use of private vehicles (Ewing, 1994, 1997, Ewing,
Pendall and Chen, 2002), or as Al Gore put it, ''A gallon of gas
canbeusedupjustdrivingtogetagallonofmilk.'' 6 Accessibility
can be quantified by measuring road length, road areas, and the
traveling times of households (Hadly, 2000).
Landscape ecology metrics can also be used to analyze acces-
sibility. For example, the size and distribution of residential
''patches'' relative to other land uses may provide a proxy mea-
sure for accessibility between these patches and commercial and
industrial ''patches.'' Accessibility can also be assessed by cal-
culating the fractal dimensions of road networks (Benguigui,
1998). Further, some ecological measures are useful to measure
accessibility, such as ''mean proximity index'' (MPI) (Gustafson,
1998; Torrens and Alberti, 2000). Another group of accessi-
bility measures is used in transportation models, including: the
isochrones measurements through which one counts the number
6 QuotefromaspeechbyAlGoreduringhiscampaign for the US presidency, January
1999. Available from http://www.greenclips.com/00issues/139.htm (accessed 15
November 2010).
Search WWH ::




Custom Search