Human activities significantly alter the geomorphology, ecology, and
climatology of the Earth at local, regional, and global scales. Nowhere is this
more apparent than in cities, which can be thought of as human-dominated
ecosystems. Cities alter surface and subsurface hydrologic and biogeochemical
processes by replacing pervious with impervious materials; change local to
regional climate by altering surface energy balances and releasing various
pollutants into the atmosphere; and influence biodiversity by fragmenting or
The expansion of cities due to population growth and migration from rural to
urban areas increasingly exposes humans to a variety of obvious (volcanoes,
earthquakes, hurricanes) and more subtle (fugitive dust, subsidence, slope
failures) geohazards. Use of remotely sensed data is frequently the only
cost-effective and timely means to characterize and assess ecological,
geological, and climatic changes resulting from urban expansion or
redevelopment. The ASTER sensor provides high to moderately high resolution data
in three wavelength regions useful for investigation of a wide range of urban
Urban Land Cover and Spatial Structure,
The high spatial resolution of ASTER in the visible to near infrared bands
(15 m/pixel) allows for detailed land cover classification of urban and peri-urban
regions. Land cover classification of urban regions is useful for a range of
applications including urban growth change detection, hydrology studies, and as
input into climate models. Figure 1 is a land cover classification of the
eastern portion of the Phoenix, AZ metropolitan region, and illustrates the
spatial and class detail extractable from ASTER data.
Figure 1. Land cover classification of eastern Phoenix metropolitan area.
Land cover data can be further analysed using landscape metrics to obtain
quantitative measures of urban spatial structure useful for ecological,
climatic, and demographic studies. There are a large number of metrics available
for urban analysis; some examples include class area, mean patch size (a measure
of the clump size of similarly classified pixels), edge density (a measure of
patch or class neighborhood shape complexity), and interspersion/juxtaposition
index (how dispersed or clumped together a class is on the landscape). Figure
2 depicts the interspersion/juxtaposition index (IJI) for the Built land
cover classes in the Phoenix urban core area, and illustrates the relatively
high degree of mixing with other land cover classes in the region.
Figure 2. Interspersion/Juxtaposition Index calculated from Phoenix land
Measurement of Surficial Biogeophysical
The broad wavelength coverage of ASTER allows for measurement of important
biogeophysical variables in urban/peri-urban regions such as vegetation density.
Using simple vegetation indices, high spatial resolution urban vegetation maps
can be made rapidly. This information is immediately useful to ecologists and
city planners for assessment of urban park extent and health. Vegetion density
is also important for modeling of urban climate, hydrology, and water use.
Figure 3 is a color ramped Normalized Difference Vegetation Index (NDVI) of
downtown London; regions with an NDVI value approaching 1 have high density of
actively photosynthesizing vegetation, and regions with values approaching -1
have little to no vegetation. The ASTER visible-near infrared (VNIR) data used to
calculate the NDVI is included for comparison.
Figure 3. NDVI and ASTER vnir data for downtown London metro area.
Collection of multispectral thermal infrared data is a particular strength of
ASTER. Nighttime data acquisitions over urban regions can be used to create maps
of urban/peri-urban surface temperature that are invaluable for assessment of
urban heat islands. The distribution of built materials throughout the urban
landscape are of obvious importance in constructing thermal budgets, but
consideration of the potential contributions of surrounding natural materials to
the regional thermal budget is also important. For example, the Phoenix, AZ
metropolitan region is bounded by mountain ranges with little vegetation cover;
these ranges act as large thermal emitters during the night and have surface
temperatures equivalent to urban core asphalt and concrete (Figure 4). This
surface temperature information is valuable for investigation of urban climatic
patterns and initialization of climate models.
Figure 4. Surface temperature map of the Phoenix metropolitan area. North
is to top of image.