Image Processing Reference
In-Depth Information
12.1
Introduction
12.1.1
Why Study Cities?
Urbanization is a significant, and perhaps the most visible, anthropogenic force on
earth - affecting its surface, atmosphere, and seas; its biodiversity and its people.
Reliable baseline data on the state of many urban area's
ecosystems and biodiversity is lacking, and our progress
in obtaining these data is moving slower than our ability
to alter the environment. Characterization and monitoring
of urban center land cover/land use change is only of limited
use in understanding the development pathways of cities
and their resilience to outside stressors (Longley
2002
).
Geological, ecological, climatic and social/political data
are also necessary to describe the developmental history of
a given urban center and to understand its ecological functioning (Grimm et al.
2000
). The data available from the NASA Earth Observing System (EOS) satellite-
based instruments presents an opportunity to collect this information relevant to
urban (areas of high population concentration with high building density and infra-
structure) and peri-urban (adjacent agricultural and undisturbed regions with low
population concentration) environments at a variety of spatial, temporal and spectral
scales. EOS sensors offer two advantages essential for characterization and moni-
toring of urban/peri-urban regions: (1) they can supply a large volume of surficial
multispectral data at relatively low or no cost, and (2) data for the same region can
be repeatedly acquired over relatively short periods (days to weeks).
urban centers are
the logical
starting point for
study of the
effects of humans
on ecosystems
and climate
12.1.2
Remote Sensing and Urban Analysis
There is a long legacy of urban and peri-urban analysis using automated, passive
satellite-based sensors, however much of this work has focused on delineation of
urban vs. nonurban land cover at coarse to moderate spatial resolutions (Donnay
et al.
2001
; Longley
2002
; Mesev
2003
). Extensive use has been made of the
Landsat Multispectral Scanner (MSS), Thematic Mapper (TM), and Enhanced
Thematic Mapper Plus (ETM+) sensors to characterize urban extent and materials
(Buyantuyev et al.
2007
; Forster
1980
; Jackson et al.
1980
; Jensen
1981
; Haack
1983
; Haack et al.
1987
; Seto et al.
2007
; Stefanov et al.
2001b, 2003
) and to con-
duct basic comparisons between urban centers (Ridd
1995
; Ridd and Liu
1998
;
Chapter 6 in this topic). As presented by Fugate et al. in Chapter 7 of this topic,
these sensors provide coarse to moderately high spatial resolution (80-15 m/pixel
in the visible and near-infrared wavelengths); fairly low spectral resolution (four to
seven bands in the visible through shortwave infrared and 1-2 thermal infrared
bands); and excellent temporal resolution (typically 14-16 day repeat cycle from
1972 to present). Other satellite-based sensors with greatly improved spatial resolution
(15 m/pixel to less than 1 m/pixel) have been developed primarily by the commercial
