Geography Reference
In-Depth Information
Table 3.7
(continued)
Code
Type
Descriptions
16
Barren or sparsely
vegetated
Land of exposed soil, sand, rocks or snow and never has more
than 10 % vegetated cover during any time of the year.
17
Water bodies
Oceans, seas, lakes, reservoirs, and rivers. Can be either fresh or
salt water.
popular and powerful one (Ghose et al.
2010
). The C4.5 classifier is selected in this
study according to the accuracy assessment to identify the vegetation disaggre-
gation classification in the farming-pastoral ecotone of north China.
The ecotones are recognized as one of the most important objects of the eco-
logical research, since they are unstable and very sensitive to the surrounding
environment. Besides, ecotones are more suitable for the study of the land cover
mapping for the climate simulation. The farming-pastoral ecotone has received a
lot of attention from academic community due to its largest area, longest span and
typical characteristics. It involves nine provinces and 106 counties, with a total
area of 654, 564 km
2
(Liu et al.
2007
). The total population in this area is
3.14 9 10
7
, with the average population density of 47.9 persons per square kilo-
meter. The land use has changed drastically throughout farming-pastoral ecotone
of North China after the widespread and profound economic reform that was
initiated in the early 1980s (Dong and Xu
2009
), and the current ratio of the
cropland, forest land and grassland is 1.0: 1.17: 3.67 (Fig.
3.9
). The temperature
rise has been more and more obvious in the past 50 years, with an average increase
rate of 0.4 C/10a (Friedl et al.
2002
).
3.3.1 Data Preparation
This study presents an inference rule of spatial data mining to distinguish forest
types based on the consistent grids in the data of the IGBP Data and Information
System (IGBPDIS, Bartholomé and Belward
2005
),
1
GLC (GLC2000, Ran et al.
2012
),
2
multi-source Integrated Chinese Land Cover (WESTDC, Liu et al.
2005
)
3
and UMD
4
land cover data in 2000. The classification rule is first rectified so as to
improve accuracy in 2000. Then the land use data of 1988, 1995 and 2005 were
converted to the land cover data according to this inference rule.
In this study, we also used the land use database developed by the CAS. The
data are available during four periods, i.e., year 1988, year 1995, year 2000 and
1
For more imformation:
http://edcdaac.usgs.gov/modis/mod12q1v4.asp
.
2
For more imformation: (
http://www.gem.jrc.it/glc2000
).
3
For more imformation: (
http://westdc.westgis.ac.cn
).
4
For more imformation: (
http://www.landcover.org
).
