Remote Sensing Analysis of Urban Sprawl in Birmingham, Alabama: 19882004'

1
Remote Sensing Analysis of Urban Sprawl in
Birmingham, Alabama 1988-2004.
Introduction In the realm of urban
studies, urban sprawl is a topic drawing a lot of
interest. Push factors of the average American
inner-city such as the perception of high crime
rates, poverty, low quality schools, and
pollution have driven the more affluent of the
urban society to seek new opportunities on the
fringe of the urban landscape. As more citizens
leave the densely constructed city center for the
more spacious suburban life, rural lands are
built-up at an accelerated rate. The
metropolitan area of Birmingham, Alabama is no
exception to this trend. Birmingham is
experiencing a constant decline of the inner-city
and expansion of suburban areas into the
surrounding rural landscape. One can see
evidence of urban expansion by driving along
roadways in the Birmingham suburbs or observing
census data for the communities and counties
surrounding the city center. According to the U.
S. Census Bureau, the population in Jefferson and
Shelby Counties increased by 10.4 between 1990
and 2005. Of the two counties, the most rapid
growth in that fifteen year period took place in
Shelby County, which experienced a population
increase of about 73 (Table 1). Table 1.
Population by County
2005 2000
1990 Jefferson County
657,229 662,047
651,525 Shelby County
171,465 143,293
99,358 Source U.S. Census Bureau, 2005
Population Estimates, Census 2000, 1990 Census
Unfortunately, census data lack the
capability to provide researchers with a detailed
landscape scale analysis of the urban expansion
taking place and the areas effected. Satellite
imagery, however, provides a birds eye view of
the physical landscape of the area of interest.
Using remote sensing and change detection
methods, the transition of areas surrounding
Birmingham from rural to urban can be seen in
great detail as the growing man-made environment
creeps its way across the mostly forested rural
landscape. Satellite remote sensing is a
relatively new source for data acquisition. The
first land sensing satellite, Landsat 1, was
launched into orbit in the summer of 1972
(Saunders et al., 2003). Since then, the Landsat
Program has archived imagery nearly twice per
month for public use. The USGS currently
maintains and collects data from Landsat 5 and
Landsat 7 (Saunders et al., 2003). For landscape
scale studies using such data is cost efficient
and combined with frequent data collection it is
a powerful tool for spatial analysis (Ryznar and
Wagner, 2001). Remotely sensed data coupled with
powerful Geographic Information Systems (GIS)
provides a tool that enables urban growth to be
monitored (Sultana and Marzen, 2004).
Results and Analysis
Data and Methods
For this project,
two cloud free Landsat 5 TM images of Jefferson
and Shelby Counties were analyzed using change
detection procedures (figure 1). For the change
detection aspect of the project, the images
selected were created 16 years apart, one from
October 13th, 1988, and the other from November
10th, 2004. The two images were examined in
tandem in order to monitor the expansion of the
urban area over the 16 year period. Both
images were classified using an ISODATA
unsupervised classification procedure. The
Landsat images were initially classified into 100
classes, each of which was designated as either
built-up (urban impervious surfaces) or not
built-up (vegetation, water, bare land) (figure
2). Next, the two images were recoded, creating
an image for each year showing the land cover as
having either a built-up or not built-up
classification. Once recoded, a
post-classification matrix process was used to
produce a final map merging and comparing the two
recoded images, allowing for analysis of
from-and-to land cover changes over the 16 year
period (figure 3).
Figure 3. Post-Classification Land Cover
Comparison
Figure 1. Landsat 5 Images of Jefferson and
Shelby Counties, Years 1988 and 2004
Table 2. Analysis Results by Class
Classes Area (Square Miles)
Percent () of Total Landscape 1. Built-up
(unchanged) 103.973
5.7 2. Built-up to not built-up
117.324 6.4 3. Not
built-up (unchanged) 1470.55
80.5 4. Not built-up to
built-up 134.511
7.4

Total
1826.36
100 The results show there
was a substantial amount urban expansion,
especially to the south and east of Birmingham
along the major highways. This analysis
indicates that remote sensing can be a powerful
tool in studying urban sprawl. However, for a
more complete analysis, future studies should not
rule out field work. In addition, other remote
sensing tools such as aerial photography could
prove to be useful. References Ryznar, R. M.
and Wagner, T. W. 2001. Using Remotely Sensed
Imagery to Detect Urban Change Viewing Detroit
from Space. Journal of the American Planning
Association, 67 (3) 327-33 Sultana, S. and
Marzen, L. 2004. Quantifying Urban Encroachment
in Dhaka, Bangladesh from 1989-2000. Theodore,
S., Feuquay, J., Kelmelis, J. 2003. The U. S.
Geological Survey Land Remote Sensing Program.
Cartography and Geographic Information Science.
30.2 (April 2003) p211(5). U. S. Census Bureau,
2006. Website www.census.gov
Figure 2. Recoded Images for 1988 and 2004