Environmental Engineering Reference
In-Depth Information
11.3
The temporal lag
challenge
11.4
Structural
-
functional
links
Urban areas are routinely represented by mixtures of land cover
and land use, where land cover types are interpreted from
remotely sensed data and land use characteristics from ground-
based surveys such as population censuses, housing, and planning
maps. The unique spatial configuration of urban land cover
types such as vegetation, soil, water and impermeable surfaces
determines the physical structure of an urban area, while human
occupation and the functioning of social and economic activities
are all indications of urban land use. When shaping spatial models
of any given urban area through the combination of land cover
types and land use characteristics, the conventional view is to
select remotely sensed data and surveys that have been taken at
similar points in time. This is the intuitive approach and the
basis for many remote sensing and general geospatial integrative
methodologies and applications.
However, this assumption of strict time-dependence may be
theoretically fallible. When consulting the established literature
on theoretical urban geography, the time relationship between
physical structure and societal functioning is generally regarded
as anything but instantaneous; and that urban structures are
an eventual consequence of functional manifestations taken and
decided upon years or sometimes decades earlier (see Whitehand,
1977; Brenner, 2000; Longley, 2002). In other words, there is an
inherent temporal lag between the reasons why society decides
to build (urban function) and when it is built (urban struc-
ture). This lag is the basis to much of the theory on urban life
cycles, urban sociological interactions, and even municipal engi-
neering (see Herbert and Thomas, 1982; Carter, 1985; Clark,
2008).
There is also evidence to suggest that, because cities are
closed systems, temporal causality is bidirectional; that struc-
ture sometimes affects function. For instance, when the building
of dense apartment complexes in one point in time some-
times leads to higher levels of poverty and even crime rates in
subsequent years, or alternatively when a mixture of condo-
miniums and green space promotes gentrification. Conventional
urban theory would further suggest that, if measured accurately,
these so-called temporal lags would be appropriate indica-
tors for measuring the pace of change in urban processes. In
particular, changes in actual physical growth resulting from
construction work, as well as the possibility to examine less
obvious indications of change such as social and economic
deprivation levels, severity of cultural segregation, exacerba-
tion of housing overcrowding, and traffic congestion. Any
insights, in either direction, of how one or more of these
urban processes fluctuate would be considered essential for
understanding the rapid dynamics of urban morphologies,
and in turn provide valuable information for the pursuit of
urban planning, legislative zoning and environmental sustain-
ability policies, both for today's cities and for their long-term
future.
Before investigating evidence for a temporal lag, it's important
to establish the relationship between structure and function at the
static scale of analysis; in particular how structural configuration
can be linked with functional characteristics.
Research in urban remote sensing has been generally confined
to measuring impermeable surfaces by classifying image pix-
els into urban built land cover. Notable breakthroughs have
augmented remotely sensed with socioeconomic information
to generate models of urban function as well as structure
(Chen, 2002; Harvey, 2002; Barnsley, Steel and Barr, 2003).
Figure 11.1 represents three types of datasets that are frequently
used for structural - functional models; high spatial resolution
sensor images to measure structure, and point-based mailing
addresses and rasterized area-based census surfaces to tessellate
socioeconomic characteristics of urban areas. Each of the three
types represents the study site of the city of Belfast, Northern
Ireland. The high spatial resolution image is from the IKONOS
sensor (Space Imaging) at 4 m, pan-sharpened and taken in July
2001, the point-based mailing addresses are from the COMPAS
database from the Ordnance Survey of Northern Ireland, and the
surface is of the 2001 Census Population is rasterized at a 200 m
grid (see Martin, Langford and Tate, 2000 for calculations).
One approach to combining structure and function, as repre-
sented by Fig. 11.1, is to perform a direct spatial comparison or
formulate a linear statistical relationship between the IKONOS
image and point-based addresses and area-based census sur-
faces at variant time slices. Commendable research includes
the estimation of population and housing units by Lo (2003)
and by Harvey (2002) by limiting census attributes within the
spatial boundaries of built land cover extracted from remotely
sensed data. In terms of point-based address data, work by
Aubrecht
et al
. (2009) and Mesev (2005, 2007) attempted to
replicate the spatial configuration of urban neighborhoods. The
assumption is that urban neighborhoods exhibit distinctive spa-
tial expressions in terms of their architectural, structural, and
morphological composition - the complex assemblage of differ-
ent land covers (bare soil, concrete, tarmac, grass, water etc.).
By employing spatial metrics to quantify these attributes it is
possible to demonstrate how individual urban neighborhoods
may be distinguished and delineated from second order imagery
(Pasaresi and Bianchin, 2001; Herold, Scepan and Clarke, 2002;
Barnsley, Steel and Barr, 2003).
On-going research is exploring an agenda for building disag-
gregated urban models that infer spatial urban syntactic structural
and functional configurations within vector-determined spectral
limitations using high spatial resolution IKONOS imagery. Dis-
aggregated models can be built either from point-based address
data extracted from COMPAS in Northern Ireland and the
United Kingdom (postal records), or from area-based parcel
data in the United States. Knowing the spatial distribution of
these point data introduces a number of key indicators that mea-
sure parameters such as density (compactness versus sparseness)
and arrangement (linearity versus randomness) (Mesev, 2007).
Commercial neighborhoods exhibit different levels of complex-
ity and irregularity to residential neighborhoods, so too does
high density residential from low density residential. This can
be measured by even the most elementary metrics, such as area,
density, and percent land cover. Fractal geometry is well suited
to measuring the structural irregularity of the morphology of
Search WWH ::
Custom Search