What is urban remote sensing? - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 5

obtain knowledge beyond our human visual perception. For
example, thermal remote sensing can measure spatially continu-
ous surface temperature that is useful to examine the urban heat
island effect (e.g., Lo, Quattrochi and Luvall, 1997). Data fusion
from different sensors can improve urban mapping and analysis
(see Ch. 10).
Third, remote sensing allows retrospective viewing of Earth's
surface, and time-series of remote sensor data can be used to
develop a historical perspective of an urban attribute or process,
which can help examine significant human or natural processes
that act over a long time period. Examples in this volume
include time-series land use/cover data that have been used to
examine the suburbanizing process in the Atlanta metropolitan
area over nearly the past four decades (Ch. 2); increasing gross
primary production (GPP) that may be linked with vegetation
carbonsequestrationduetourbangrowthintheeasternUnited
States (Ch. 19); historical land use changes affecting upon near-
surface air temperature during recent extreme heat events in
the Phoenix metropolitan area (Ch. 21); and urban growth and
landscape changes affecting biodiversity in northern Washington
(Ch. 25).
Fourth, remote sensing can help make connections across
levels of analysis for urban studies. Urban science disciplines and
subdisciplines have their own preferred levels of analysis and
normally do not communicate across these levels. For example,
urban planners tend to work at street and neighborhood levels;
regional planners deal with a larger environment such as several
counties, one or more metropolitan areas, or even a whole
state; urban meteorologists and ecologists tend to work at levels
defined by physiographical features or ecological units; and
urban geographers tend to work at various levels depending
upon specific topics under investigation. On the other hand, the
temporal scales used by these different urban researchers vary
greatly, from hourly, daily, weekly, monthly, seasonally, to annual
or decadal basis. Remote sensing provides essentially global
coverage of data with individual pixels ranging from submeters
to a few kilometers and with varying temporal resolutions; such
data can be combined to allow work at any scales or levels
of analysis, appropriate to the urban phenomenon or process
being examined. Therefore, remote sensing offers the potential
for promoting urban researchers to think across levels of analysis
and to develop theories and models to link these levels.
Last, remote sensing integrated with relevant geospatial tech-
nologies, such as geographic information systems, spatial analysis
and dynamic modeling, offers an indispensible framework of
monitoring, synthesis and modeling in the urban environment.
Such frameworks support the development of a spatio-temporal
perspective of urban processes or phenomena across different
scales and the extension of historical and current observations
into the future. They can also be used to relate different human
and natural variables for developing an understanding of the
indirect and direct drivers of urban changes and the poten-
tial feedbacks of such changes on the drivers in the urban
environment.
Nevertheless, urban environments are complex by nature,
challenging the applicability and robustness of remote sensing.
The presence of complex urban impervious materials, along
with a variety of croplands, grasslands and vegetation cover,
causes substantial interpixel and intrapixel scenic changes, thus
complicating the classification and characterization of urban
landscape types. Moreover, it is always difficult to integrate
remote sensor data with other types of geospatial data in urban

social or environmental analyses because of the fundamental
differences in data sampling and measurement. Some additional
challenges will be addressed in the sections to be followed.

1.3 Remote sensing
systems for urban areas

Remote sensor data used for urban studies should meet cer-
tain conditions in terms of spatial, spectral, radiometric, and
temporal characteristics (Jensen and Cowen, 1999). There is a
wide variety of passive and active remote sensing systems acquir-
ing data with various resolutions that can be useful for urban
studies. Medium-resolution remote sensor data have been used
to examine large-dimensional urban phenomena or processes
since early 1970s when Landsat-1 was successfully launched.
With the launch of IKONOS, the world's first commercial,
high-resolution imaging satellite, on September 24 1999, very-
high-spatial-resolution satellite imagery became available, which
allow detailed work concerning the urban environment. Inde-
pendent of weather conditions, active remote sensing systems,
such as airborne or space-borne radar, can be particularly useful
for such applications as housing damage assessment or ground
deformation estimation in connection to some disastrous events
in urban areas. Another active sensor system, similar in some
respects to radar, is lidar (light detection and ranging), which can
be used to derive height information useful for reconstructing
three-dimensional city models.
With five major chapters, Part II of this volume reviews
some major advances in remote sensors that are particularly
relevant for urban studies. It begins with a chapter (Ch. 2)
discussing the utilities of medium-resolution satellite remote
sensing for the observation and measurement of urban growth
and landscape changes, emphasizing the use of the data from the
Landsat imaging sensors. Over a period of nearly four decades,
the Landsat program has acquired a scientifically valuable image
archive unmatched in quality, details, coverage, and length, which
has been the primary source of data for urbanization studies at
the regional, national and global scales. The chapter comprises a
moderate review on the past, present and future of the Landsat
program and its imaging sensors, a case study focusing on a
rapidly suburbanizing metropolis, and an extended discussion
on some conceptual and technical issues emerging when using
archival satellite images acquired by different sensors and perhaps
during different seasons.
The other four chapters within Part II review the utilities
of high-resolution optical and radar remote sensing, hyperspec-
tral remote sensing, and lidar remote sensing for urban feature
extraction. Chapter 3 discusses some major challenges and limi-
tations when using very-high-resolution optical satellite imagery
for monitoring human settlements, including geometric, spec-
tral, classification, and change detection problems. Then, the
authors propose an integrated spatial approach to deal with some
of these problems, which is followed by a discussion of some
interesting results using very-high-resolution satellite imagery
for building damage assessment in connection to major earth-
quake events. Chapter 4 reviews the methodological development
of urban hyperspectral remote sensing emphasizing the progress
in developing an automated system for mapping urban surface
materials. This system comprises an iterative procedure that