Image Processing Reference
In-Depth Information
measures in the context of urban places is the contiguity index, which is a measure
of the “clumpiness” of land cover classes. In particular, we are interested in the
clumpiness of impervious surfaces because we hypothesize that high levels of
clumpiness (where pixels of the impervious surface land cover class are all in close
proximity to one another) represent one or only a few buildings, characteristic of
central cities and other dense areas. On the other hand, low levels of clumpiness
of impervious surface should represent a disaggregation of pixels of the same land
cover class, representing a greater number of buildings, associated with lower
density, more suburban areas.
Two areas might have identical fractions of impervious surface, but the one with
a high contiguity index would probably represent a “more urban” area than the one
with the lower contiguity index. In general, we would expect that city centers would
have the highest abundance of impervious surface and also the highest level of
contiguity of that impervious surface. At the other extreme, a place that is not very
urban will have a low proportion of impervious surface, but that surface might be
highly contiguous (one small building) or only moderately so (three small build-
ings), but the degree of contiguity would matter less than it would when the propor-
tion of impervious surface is high. This suggests that the configuration of the pixels
increases in importance as the proportional abundance of impervious surface
increases, implying the existence of an exponential relationship.
The way in which these several measures of composition and configuration can
be most satisfactorily put together is still under investigation (see Weeks
2004
;
Weeks et al.
2005
). However, the research conducted thus far suggests the utility of
this approach to the creation of an urban index that can be combined with census
data to characterize the nature of urbanness of a place.
3.4
Using the Urban Index as a Predictor Variable
An urban index of the type that I have suggested may be of inherent interest on its
own, but its greatest value in social science research is almost certainly that it pro-
vides a way of contextualizing the environments in which people live. Places that
are different in terms of urbanness are likely to be different in other ways that will
affect the lives of the people there. Similarly, changes over time in urbanness can be
expected to be related, both causally and consequentially, to the lives of the people
who comprise the residents and/or workers in those changing environments.
As long as the researcher is careful to use the same measurements from the satel-
lite imagery and census data over space and time, then differences in the urban
index can be proxies for differences between places and changes over time in the
social and economic aspects of the people being studied. This characteristic of a
place can than be introduced into a regression analysis as a predictor variable, or
even into multi-level analyses as a community-level factor that may be related
to individual behavior taking place in different places and/or at the same place at
different times.
