International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
lable 3. Statistics of simulation results for different development scenarios.
2010 2030 2050 during 2010-2030 are largely
Area Percent Area Dercent Arca Perceni distributed over places far away
(hectare) (hectare) (hectare) from the 1999 urban land.
2 50-59** 41985 257 21767 1.33 12390 0.76 Many projected additions are
SS 60-69 51817 3.17 29716 1.82 17447 1.07] also found in western,
$3 70-79 60651 3.71 43419| 2.65 26173 eon Lol ne muster, and southeastern,
5| 5$ 80-89 61269] ^ 375 76326] 467 cs Sq. D Some US unen
Tom Fun EEE i a
: : i Lo = Pe recognized. The projected
l'otal urban area* ** 736395 45.03 1133787 69.32 1286692 78.67 urban additions after 2030 are
50-59 29843 1.82 31444 1.92 22239 1.36 predominately scattered over
9 = 60-69 22038 1.35 41307 2.54 31110 1.0( the western and southeastern
e 2 70-79 7188 0.44 54950 3.36 45942 2.81. parts. Under the second
2 80-89 559 0.03 75635 4.62 74920 4.58 scenario, the projected
5 m 90-100 0 0.00 75848 4.64 238798 14.6 urbanization for the period of
^| "Total urban area 352758. 3379... 7725140 47.3 906134] — 55.4 1999-2010 has been very
* It is computed by using urban area divided by the total modeled area (1.635.656 hectares). limited. Mos! of he DN
d l'his is the probability of predicted urbanization. additions are for the period of
men 2010-2030, represented by blue
slope steepness for urban land would decrease from 4.87 percent
in 1999 to 4.46 percent in 2050, implying that many crop land
It contains 1999 urban area (493,131 hectares)
would be converted into urban uses.
projected urban additions
and green pixels in Figure 2
(Note: The original figure is in
color. Readers can contact the author for obtaining a copy of this
color plate). Numerous large urban clusters can be clearly
recognized, particularly in southern and western parts.
The spatial distribution of simulated urbanization under different
scenarios can be discerned from Figure 2. For the first scenario,
the projected urban additions for the period of 1999-2010 are
largely adjacent to the 1999 urban pixels, which can be viewed
as ‘continuation’ of urbanization. This is in line with the statistics
given in Table 3, which show that more than 99 percent of the net
urban growth under this scenario are accounted for by the organic
This type of urban growth actually represents the
expansion of existing urban pixels into their surroundings. The
growth.
4. CONCLUSION
Scenario One
Scenario Two
2010
2030
Probability of predicted urbanization
EE 90 - 100 percent [7] 70 ~ 89 percent [E] 50 — 69 percent
Others
[Urban extent in 1999
Water [4 Hillshaded relief
2050
Figure 2. Simulation of the spatial consequences of future urban growth under different
scenarios. Note that the county boundary is overlaid.
1230
This study has demonstrated the usefulness of remote sensing,
dynamic modeling and geographic information technologies for
urban planning. The model used here has been tested by its
developers for long-term urban growth prediction in two study
sites. From a user's perspective, this study has moved forward to
investigate the effectiveness of the model as a tool to imagine, test
and choose between different
scenarios. These scenarios
represent different growth strategies
that can be adopted by planners.
This is an area on which substantial
research efforts need to be made in
order to adopt dynamic modeling
technology for problem solving in
applied urban studies.
At the application level, this study
has established a well-documented
regional case study focusing on
Atlanta, a metropolis without any
major natural barriers. The two
scenarios designed with different
environmental and policy conditions
have largely represented the major
possible planning strategies. These
result from the first scenario
indicates Atlanta would be the next
‘Los Angeles' by approximate 2030
if the current rate and pattern of
urban growth do not alter. These
will serve as a good warning to
planners in Atlanta. In contrast, the
result from the second scenario
shows that much more greenness
and open space, including buffer
zones of large streams and lakes,
could be preserved. Accordingly,
the second scenario should be the
most desirable for the future urban
