4.4.2 Change Dynamics Analysis
In terms of coarsely multi-temporal change detection method, intermediate LULC
classification maps have been generated with given time interval and are used and
analyzed to reveal the change processes during the time period. However, to detect
more complex dynamics of LULC changes, all classification maps need to be
involved to accomplish time-serial analysis. A trajectory of change process of
each class is clearly illustrated in Fig. 4.8. It shows a dynamic change of each
class too.
Obviously, urban built-up area experiences growth over the 30 years in general.
However, based on the statistics recorded in Table 4.6, the overall trend of built-up
area has been increased, but the process is torturous. The area increases in 1 year,
but falls back slightly in next year and increases again in later years. For example,
according to the table total percentage of built-up area is 48.3 % in 2002, but the
value decreases to 47.9 % in 2003 and rebounds to 51.3 % in 2005. The situation of
vegetated area is similar as built-up area that the coverage decreases over the period
but experiences fluctuation. As for forest land and barren land, they have erratic
change throughout the years, while water cover keeps relatively stable values
around 2 %. Based on the observation of built-up area over the entire time span,
the Region of Waterloo has experienced relatively accelerating urbanization process in 1990s and in early 2000s.
Area net change
1984~1990 1990~1996 1996~2002 2002~2008 2008~2013
Time period
(%)
15
10
5
Water
Forest land
Vegetated area
Low-density urban area
High-density urban area
Barren land
0
-5
-10
-15
Fig. 4.7 Normalized net change in municipality area by time period for each LULC class
80
A. Fu et al.
In terms of coarsely multi-temporal change detection method, intermediate LULC
classification maps have been generated with given time interval and are used and
analyzed to reveal the change processes during the time period. However, to detect
more complex dynamics of LULC changes, all classification maps need to be
involved to accomplish time-serial analysis. A trajectory of change process of
each class is clearly illustrated in Fig. 4.8. It shows a dynamic change of each
class too.
Obviously, urban built-up area experiences growth over the 30 years in general.
However, based on the statistics recorded in Table 4.6, the overall trend of built-up
area has been increased, but the process is torturous. The area increases in 1 year,
but falls back slightly in next year and increases again in later years. For example,
according to the table total percentage of built-up area is 48.3 % in 2002, but the
value decreases to 47.9 % in 2003 and rebounds to 51.3 % in 2005. The situation of
vegetated area is similar as built-up area that the coverage decreases over the period
but experiences fluctuation. As for forest land and barren land, they have erratic
change throughout the years, while water cover keeps relatively stable values
around 2 %. Based on the observation of built-up area over the entire time span,
the Region of Waterloo has experienced relatively accelerating urbanization process in 1990s and in early 2000s.
Area net change
1984~1990 1990~1996 1996~2002 2002~2008 2008~2013
Time period
(%)
15
10
5
Water
Forest land
Vegetated area
Low-density urban area
High-density urban area
Barren land
0
-5
-10
-15
Fig. 4.7 Normalized net change in municipality area by time period for each LULC class
80
A. Fu et al.