Download Surface modelling of human population distribution in China

Ecological Modelling 181 (2005) 461–478
Surface modelling of human population distribution in China
Tian Xiang Yuea,∗ , Ying An Wanga , Ji Yuan Liua , Shu Peng Chena , Dong Sheng Qiua ,
Xiang Zheng Denga , Ming Liang Liua , Yong Zhong Tiana , Bian Ping Sub
a
Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences,
917 Building, Datun, Anwai, Beijing 100101, China
b College of Science, Xi’an University of Architecture and Technology, Xi’an 710055, China
Received 24 March 2003; received in revised form 23 April 2004; accepted 4 June 2004
Abstract
On the basis of introducing major data layers corresponding to net primary productivity (NPP), elevation, city distribution
and transport infrastructure distribution of China, surface modelling of population distribution (SMPD) is conducted by means
of grid generation method. A search radius of 200 km is defined in the process of generating each grid cell. SMPD not only pays
attention to the situation of relative elements at the site of generating grid cell itself but also calculates contributions of other
grid cells by searching the surrounding environment of the generating grid cell. Human population distribution trend since 1930
in China is analysed. The results show that human population distribution in China has a slanting trend from the eastern region
to the western and middle regions of China during the period from 1930 to 2000. Two scenarios in 2015 are developed under
two kinds of assumptions. Both scenarios show that the trends of population floating from the western and middle regions to the
eastern region of China are very outstanding with urbanization and transport development.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Surface modelling; Population distribution; Grid generation; Geographical information system
1. Introduction
A surface model is the mathematical representation of a surface in such a form that it can be used
in design calculations. Since the first digital terrain
model for road design was produced by the Massachusetts Institute of Technology in 1957, surface
modelling has begun to be developed (Stott, 1977; Yue
∗ Corresponding author. Tel.: +86 10 64889633;
fax: +86 10 64889630.
E-mail address: [email protected] (T.X. Yue).
0304-3800/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.ecolmodel.2004.06.042
et al., 2004). Surface modelling includes development
of digital terrain models, spatial interpolation models,
area-based matching models and a multi-resolution approach. Three widely used principal ways of structuring a digital terrain model are triangulated irregular networks, regular grid networks and contour-based
networks (Moore et al., 1992). Spatial interpolation
models include interpolation by drawing boundaries,
trend surface analysis, moving averages, Kriging interpolation, spline curves and finite element method
(Stein, 1999; Sabin, 1990; Shipley, 1990). Area-based
matching models include radiometric model and least
462
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
squares correlation (Mustaffar and Mitchell, 2001;
Heipke, 1997; Gruen, 1985; Foestner, 1982). The
multi-resolution approach is an image-driven surface
estimation, which is characterized by three phases that
are shape modelling, multi-resolution model construction and variable resolution representations (Sarti and
Tubaro, 2002; Cignoni et al., 1998).
Surface modelling of population distribution
(SMPD) that was developed on the basis of grid generation method (Yue et al., 2003) is aimed at formulating population in a regular grid system, in which each
grid cell contains an estimate of total population that
is representative for that particular location. Compiling
population data in grid form is by no means a new approach. For instance, Adams (1968) presented a computer generated grid map of population density in west
Africa; Population Atlas of China presented grid population data for several regions in China (Institute of
Geography of Chinese Academy of Sciences, 1987);
transforming population data from census to grid
(Tobler et al., 1997), apportioning census counts to each
grid cell based on probability coefficients (Dobson et
al., 2000), and estimating population using nighttime
light data (Sutton, 1997; Sutton et al., 1997, 2001; Lo,
2001; Sutton et al., 2003).
2. Background of major data layers
In addition to total population in every province,
the major data layers that are matched with their corresponding SMPD variables include net primary productivity (NPP), elevation, city distribution and transport
infrastructure distribution. The elevation is a natural
factor and has a slow change with time so that it is
spatial variable and could be regarded as a temporal
Fig. 1. Spatial distribution of cities in 2000 in China (unit: thousand persons grouped by non-agricultural population in urban district).
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
constant within 100 years. Although NPP is based on
climate and soil, it could be modified by human activities so that it is a spatial and temporal variable. City
distribution and transport infrastructure distribution are
spatial and temporal variables, which are determined
by both natural factors and human activities, have a
rapid change with time in China during recent 100
years.
2.1. Spatial distribution of cities in China
Urbanization is a process of the concentration of
population in cities. Spatial distribution of cities and
proximity to cities are essential factors for human population distribution of China. The spatial distribution
of cities in China has had the feature that city density is much higher in eastern China than in west-
463
ern China in modern history. During the period from
1843 to 1893, urban population proportion had slow
growth, which was increased from 5.1% to 6.0% averagely in China; in the area of lower reaches of Yangtze
River the urban population proportion was increased
from 7.4% to 10.6%, in the coastal area of south
China from 7.0% to 8.7%; in inland area the proportion paced up and down between 4.0% and 5.0%.
From 1895 to 1931, in areas along coast and Yangtze
River, northeast China and north China cities were developed rapidly, while in inland area cities were developed very slowly, even at a standstill. In the early
1930s, urban population proportion was about 9.2% in
China. From 1931 to 1949, the turbulent and unstable
situation led to slow population growth in China and
the urban population proportion increased to 10.6%
(Zhang, 1997). Cities in China spatially concentrates
Fig. 2. Spatial distribution of railways in 1930 in China.
464
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
in coastal area, especially Yangtze River Delta, Peal
River Delta and Beijing–Tianjin–Tangshan area. In
2000, 42.1% of the 667 major cities of China distributed in eastern China where area accounts for 9.5%
of the whole area of China; 34% distributed in middle China where area accounts for 17.4%; 23.8% distributed in western China where area accounts for
70.4% (Urban Society and Economy Survey Team of
National Bureau of Statistics of People’s Republic of
China, 2001). The distribution densities of cities in
eastern China and in middle China were respectively
13.1 times and 5.8 times the one in western China
(as seen in Fig. 1). The urban population proportion
was 36.22% in 2000 (National Bureau of Statistics
of People’s Republic of China, 2001). According to
National Report of China Urban Development (China
Mayor Association, 2002), the urban population proportion would be 46.9% in 2015. If the urban popula-
tion proportion would increase at the average rate in
recent 5 years, annually 1.44%, it would be 57.82%
in 2015.
2.2. Spatial distribution of transport infrastructure
in China
Transport infrastructure is a primary indicator of
human population distribution (Dobson et al., 2000).
Roads and railways are especially indicative because
of their vital role in human well being. Construction
of a piece of railroad in 1881, which was from TangShan city to Feng-Nan county about 9.7 km, initiated
railroad development in China. There was 14,411 km
of railroad in China in 1930 (Fig. 2) and 21,800 km
in 1949 (Fig. 3). Length of railway in operation was
68,700 km in 2000 (Fig. 4) (Year Book House of China
Transportation and Communications, 2001). However,
Fig. 3. Spatial distribution of railways in 1949 in China.
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
465
Fig. 4. Spatial distribution of railways in 2000 in China.
relative research results showed that the appropriate
length of railroad should be 100,000 km in China, from
which current railroad length has a great gap (Chen and
Zhang, 2000). To implement the western development
strategy, railroad building in China is paying attention
to strengthening the linkage between eastern region and
western region of China, speeding up construction of
accesses to central Asia and southeast Asia and improving connection within the western region of China. Total length of railroad in China would reach 81,653 km
in 2015 (Fig. 5).
In 1902, the first automobile was imported in
China and in 1906 the first piece of road was constructed. In 1949, length of highways that automobiles could go through was 80,700 km (Fig. 6). After construction for 50 years, total length of highways was about 1.4 million km in 2000 (Fig. 7) (Year
Book House of China Transportation and Communi-
cations, 2001). In recent 10 years, major projects of
highway construction include seven east–west main
trunk roads and five south–north main trunk roads
as well as three important sections, of which total
length is about 35,000 km (Fig. 8). The seven east–west
main trunk roads include highways from Suifenhe
to Manzhouli, from Dandong to Lasa, from Qingdao to Yinchuan, from Lianyungang to Huerguosi,
from Shanghai to Chengdu, from Shanghai to Ruili
and from Hengyang to Kunming. The 5 south–north
main trunk roads include highways from Tongjiang of
Heilongjiang province to Sanya of Hainan province,
from Beijing to Fuzhou, from Beijing to Zhuhai,
from Erlianhaote to Hekou and from Chongqing to
Zhanjiang. The 3 important sections include highways
from Beijing to Shenyang, from Beijing to Shanghai and passageway going abroad from southwestern
China.
466
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
Fig. 5. Spatial distribution of railways in 2015 in China.
2.3. Land use and spatial distribution of NPP in
China
Land use is a good indicator of spatial human population distribution. In most regions, population would
range from extremely low density in desert, water, wetlands, ice or tundra land cover to high density in developed land cover associated urban land cover, between which arid grasslands, forests and cultivated
lands would range (Dobson et al., 2000; Liu et al.,
2003a). The land-use database of China during the period of 1980s and 1990s (Fig. 9), which is derived from
Landsat Thematic Mapper (TM) imagery at 30-m resolution (Liu et al., 2003b).
Net primary productivity is the difference between
accumulative photosynthesis and accumulative autotrophic respiration by green plants per unit time and
space (Lieth and Whittaker, 1975). Terrestrial Ecosys-
tem Model (Liu, 2003) is employed for analysing spatial distribution of NPP in China (Fig. 10). It integrates different data types that include the land-use
change data, daily climatic data and soil data. The analysis results show that the mean annual NPP of terrestrial ecosystems in China was 3.588 × 1015 gC year−1
in 1990s, which is greater 0.094 × 1015 gC year−1
than the one in 1980s. In other words, the mean annual NPP increased by 0.49 gC m−2 year−1 during the
20 years.
The general situation in China is that from southeast to northwest NPP becomes smaller and smaller
gradually. Most of the NPP is distributed in the East
of the rainfall line where the annual precipitation
is 410 mm, excepting that there is higher NPP in
the southern slopes of Tianshan mountains and
Altai mountains in Xinjiang. The maximum NPP
appears in Xiaoxinganling mountain and Changbai
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
467
Fig. 6. Spatial distribution of roads in 1949 in China.
mountain in the northeast China, Yunnan–Guizhou
Plateau, Guangxi, Hainan, Chongqin and provinces
along middle and lower reaches of Yangtze
River.
In terms of land-use types, on the average, NPP of
shrub and open forest is 1071 gC m−2 year−1 , evergreen broad-leaved forest 975 gC m−2 year−1 , deciduous broad-leaved forest 928 gC m−2 year−1 , coniferous
and broad-leaved mixed forest 870 gC m−2 year−1 ,
farmland system 752 gC m−2 year−1 , evergreen
coniferous forest 587 gC m−2 year−1 , deciduous
coniferous forest 585 gC m−2 year−1 and grassland
271 gC m−2 year−1 (Liu, 2001).
2.4. Elevation
Elevation is an important variable of human population distribution because most human settlements occur
on lower elevation in China. For instance, area of plains
and hills with an elevation lower than 500 m accounts
for about 28% of total land area of China, where 74%
of total population in China inhabit (Zhang, 1997). The
terrestrial parts of China are broadly divided into three
steps (Fig. 11) from Qinghai–Xizang Plateau eastward
(Zhao, 1986). The lofty and extensive Qinghai–Xizang
Plateau is the first great topographic step. Its eastern
and northern borders roughly coincide with the 3000 m
contour line. It generally has an elevation of 4000 m to
5000 m and hence is called the roof of the world.
From the eastern margin of the Qinghai–Xizang
Plateau eastward up to the DaHinggan–Taihang–
Wushan mountains lies the second great topographic
step. It is mainly composed of plateaus and basins with
elevations of 1000–2000 m, such as the Nei Mongol,
Ordos, Loess and Yunnan–Guizhou Plateaus and the
Tarim, Junggar and Sichuan basins.
468
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
Fig. 7. Spatial distribution of roads in 2000 in China.
From the eastern margin of the second step eastward
up to the coast is the third great topographic step. The
largest plains of China, the northeast China Plain, the
north China Plain and the middle and lower Changjiang
Plain are distributed in this step, which generally lie at
elevation of below 200 m.
2.5. Population growth
Since 1930, population in China has increased about
three times (Table 1). Hu’s result (1935, 1983) showed
that total population of China was 452.8 million persons in 1930 and 541.67 million persons in 1949. Both
fertility rate and death rate were higher and natural
rate of population growth was lower in the period from
1930 to 1949. During the period from 1950 to 2000, total population increased by 725.74 million persons, of
which the annual mean growth rate was 2.7%. The pop-
ulation growth underwent a rapid increase stage from
1950 to 1973, in which fertility rate was higher and
death rate was lower and a relative slow stage after
birth control policy, only one child for one couple, has
been carried out in China in 1973, in which both fertility rate and death rate were lower. Although the birth
control policy has restrained rapid population growth,
annual newborn children are still more than 9.5 million
in recent years in China because of the huge base number (Research Center for Population of CASS, 1985;
Institute of Population and Labor Economics of CASS,
2001). The projection results (Jiang, 1998), on the basis of comprehensively analysing all factors that affect
population growth in China, show that population in
China under assumptions of higher total fertility rate
and lower total fertility rate would be 1457.84 million
persons and 1417.78 million persons, respectively, in
the year 2015.
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
469
Fig. 8. Spatial distribution of roads in 2010 in China.
3. Methods and results
NPPij (t) = exp −
3.1. SMPD
By means of grid generation method (Morrison,
1962; Sidorov, 1966; Ahuja and Coons, 1968; Liseikin,
1999), the simulation model for population distribution (SMPD) is developed (Yue et al., 2003), which is
a transformation between computational domain (i,j)
and physical domain (i,j,MSPDij (t)).
pij (t)
MSPDij = G(n, t) pij (t)
pij (t) = Wij (t)(NPPij (t))0.0001 (DEMij (t))0.7

1.2
M(t)
(Sk (t))

× (Tranij (t))1.3 
dijk (t)
k=1
(1)
(2)
Tranij (t) =
DEMij (t) =
(MNPPij (t) − 760)2
106
raij (t) + roij (t)
maxi,j {raij (t) + roij (t)}
 500

demij (t) ≥ 3700 m

 (demij (t))2
500


 demij (t)
1
500 m < demij (t) < 3700 m
(3)
(4)
(5)
demij (t) ≤ 500 m
where t is a time variable; G(n,t) is total population in
province n at time t, in which grid cell (i,j) is located,
or whole China; Wij (t) is an indicative factor of water
area, when grid cell (i,j) is located in water area Wij (t)
= 0, or else Wij (t) = 1; Tranij (t) is a transport infrastructure factor of grid cell (i,j); NPPij (t) is a factor of
net primary productivity of grid cell (i,j); DEMij (t) is
470
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
Fig. 9. Land cover of China in 2000.
Fig. 10. Spatial distribution of the mean NPP in 1990s in China (unit: gC m−2 year−1 ; after Liu, 2001).
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
471
Fig. 11. Digital elevation model of China.
an elevation factor of grid cell (i,j); Sk (t) is size of the
kth city; M(t) is the total number of cities; dijk (t) is the
distance from grid cell (i,j) to the core grid cell that has
the highest population density in the kth city; raij (t)
and roij (t) represent, respectively, rail density and road
density at grid cell (i,j); MNPPij (t) is the mean annual
net primary productivity at grid cell (i,j); demij (t) is
elevation at grid cell (i,j).
3.2. Simulation process
SMPD simulates population distributions at four
times-points that are the years of 1930, 1949, 2000
and 2015. Because population data in every province
are available in 1930, 1949 and 2000, G(n,t), n = 1, 2,
. . ., 31, represent provincial population in the process
of population distribution simulation at the first three
times-points. Population projection for 2015 is only
carried out on national level so that G(n,t), n = 1, represents population of the whole China, in the process
of developing scenarios at the last times-point.
The major auxiliary tools of grid generation include
ArcInfo GIS and Delphi computer language. Nine
data layers are involved, which are NPP (net primary
productivity), LU (land use database), DEM (digital
elevation model), WA (water area), GridRail (railway
network), GridRoad (road network), Chbnd (administrative boundary), Chzh (urban area) and Cityshp
472
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
Table 1
The provincial population of China excluding Taiwan, Hong Kong and Macao temporarily
Region
Area (km2 )
Population density (person per km2 )
Population size (million persons)
1930
1949
2000
1930
1949
2000
Western China
Inner Mongolia
Guangxi
Chongqing
Sichuan
Guizhou
Yunnan
Tibet
Shaanxi
Gansu
Qinghai
Ningxia
Xinjiang
6725746
1143327
236544
82390
483759
176109
383101
1201653
205732
404622
716677
51785
1640111
110.63
4.40
11.50
Belong to Sichuan
51.34
11.03
11.52
0.78
10.39
5.49
1.28
0.39
2.51
174.57
37.88
18.42
Belong to Sichuan
57.30
14.16
15.95
1.00
13.17
9.68
1.48
1.20
4.33
354.60
23.01
47.24
30.91
84.07
36.77
40.77
2.51
35.72
25.34
4.80
5.54
17.92
16
4
49
Belong to Sichuan
106
63
30
1
50
14
2
8
2
26
33
78
Belong to Sichuan
118
80
42
1
64
24
2
23
3
53
20
200
375
174
209
106
2
174
63
7
107
11
Middle China
Shanxi
Anhui
Jiangxi
Henan
Hubei
Hunan
Jilin
Helongjiang
1670726
156563
140165
166960
165619
185950
211815
191093
452561
150.15
11.30
21.92
17.16
31.92
25.94
29.54
7.82
4.55
161.41
12.81
27.86
12.68
41.74
25.36
29.87
10.09
1.01
419.41
31.96
62.78
41.64
95.27
59.36
65.15
26.27
36.98
90
72
156
103
193
140
139
41
10
97
82
199
76
252
136
141
53
2
251
204
448
249
575
319
308
137
82
Eastern China
Beijing
Tianjin
Hebei
Liaoning
Shanghai
Jiangsu
Zhejiang
Fujian
Shandong
Guangdong
Hainan
1203528
16386
11620
188111
146316
8013
103405
103196
122468
157119
179776
40070
187.23
1.52
1.47
30.29
16.08
3.91
30.29
20.07
13.99
36.67
32.93
Belong to
Guangdong
205.68
4.14
3.99
30.86
18.31
5.06
35.12
20.83
11.88
45.49
30.00
Belong to
Guangdong
462.71
11.14
9.19
66.71
41.35
13.22
70.69
45.01
33.05
89.75
74.99
7.61
156
93
126
161
110
488
293
194
114
233
183
Belong to
Guangdong
171
253
344
164
125
632
340
202
97
290
167
Belong to
Guangdong
(geographical coordinate of city). The data are first
pre-processed as follows: (1) converting NPP into vector data, (2) overlaying Chbnd with GridRoad and
GridRail by Intersect and creating a data layer, ChBndNew, (3) adding fields, CityFlag for urban code and
rural code and CityArea for areas of urban districts, in
Chzh, (4) overlaying Chzh with ChBndNew by Intersect and creating a data layer, ChCity, (5) overlaying
NPP with ChCity by Intersect and creating a data layer,
NppNew, (6) overlaying LU with NppNew by Intersect
and creating a data layer, LNpp, (6) overlaying DEM
384
680
791
355
283
1650
684
436
270
571
417
190
with LNpp by Intersect and creating a data layer, DLNpp and (7) overlaying WA with DLNpp by Intersect
and creating a data layer, WDLNpp.
Every grid cell in 1 km × 1 km resolution is generated on the basis of WDLNpp, which includes six steps:
(1) to read the attribute values of natural and socioeconomic indicators at every grid cell, (2) to calculate
the contribution of NPP and elevation to the generating grid cell, (3) to define a search radius of 200 km
and to search cities and transport infrastructures that
have considerable effects on the generating grid cell,
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
473
Fig. 12. The human population distribution of China in 1930 (unit: persons per square kilometer).
(4) to calculate the contribution of the searched cities
and transport infrastructures to the generating grid cell,
(5) to operate the SMPD and (6) text file of the calculated result is converted into point vector data and grid
data is created from the point vector data.
3.3. Results
According to current ecological and economical situation, China could be geographically analysed in three
regions that are western, middle and eastern China. The
western region of China consists of five provinces in
southwest China, five provinces in northwest China,
Inner Mongolia Autonomous region and Guangxi
Zhuang Autonomous region. The five provinces in
southwest China are Sichuan province, Chongqing
city, Yunnan province, Guizhou province and Tibet
Autonomous region. The five provinces in northwest
China are Shaanxi province, Gansu province, Ningxia
Hui Autonomous region, Xinjiang Uygur Autonomous
region and Qinghai province. Area of the western region of China is about 6.7546 million km2 , accounting for 70% of the whole of China. The middle region
of China consists of eight provinces that are Shanxi,
Anhui, Jiangxi, Henan, Hubei, Hunan, Jilin and Helongjiang, of which area is 1.67 million km2 and accounts for 17.4% of the Whole of China. The eastern
region of China consists of 11 provinces that are Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong and Hainan, of
which area is 1.2 million km2 and accounts for 12.5%
of the whole of China.
A comparison of the simulation results shows that
the ratios of population in the western region of China
to total population of China were 24% in 1930 (Fig. 12),
32% in 1949 (Fig. 13) and 29% in 2000 (Fig. 14); the
ones in the middle region were 33% in 1930, 30%
in 1949 and 34% in 2000; the ones in the eastern
region were 41% in 1930, 38% in 1949 and 37% in
2000. Human population had a slanting trend from
the eastern region to the western and middle regions
of China during the period from 1930 to 2000. From
1930 to 1949, on an average, annual growth rate of
population was 3% in the western region, 0.4% in the
middle region and 0.5% in the eastern region; from
1949 to 2000, annual growth rates of population were
474
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
Fig. 13. The human population distribution of China in 1949 (unit: persons per square kilometer).
Fig. 14. The human population distribution of China in 2000 (unit: persons per square kilometer).
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
475
Table 2
Population change of in different regions of China excluding Taiwan, Hong Kong and Macao temporarily
Years and scenarios
The western region
The middle region
The eastern region
Population
Ratio (%)
Population
Ratio (%)
Population
Ratio (%)
1930
1949
2000
110.63
174.57
354.6
24
32
29
150.15
161.42
419.41
33
30
34
187.23
205.68
462.71
41
38
37
2015
Scenario I
Scenario II
180.86
196.55
12
14
427.56
439.31
29
31
849.42
781.92
58
55
respectively 2%, 3.1% and 2.5% in the western, middle
and eastern region (Table 2).
From now to 2015, all of urban population proportion, large-scale highway construction and population
growth have at least two possibilities in China. If the
increase of urban population proportion at the average
rate in recent 5 years, annually 1.44%, the urban population proportion would be 57.82% in 2015; but in
terms of the National Report of China Urban Development, the increase rate will be 1.0% annually and
the urban population proportion will be 46.9% (China
Mayor Association, 2002). The major highway construction projects will be completed in 2010. During the
period from 2010 to 2015, construction length of highways might be continued at the increase rate, 0.25%
annually, as in the period from 2000 to 2010; it is also
possible that large-scale highway construction stagnate
temporally. In terms of the projection (Jiang, 1998),
population would be 1457.84 million persons at higher
total fertility rate and would be 1417.78 million persons
Fig. 15. Scenario I of the human population density (unit: persons per square kilometer).
476
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
Fig. 16. Scenario II of the human population density (unit: persons per square kilometer).
at lower total fertility rate in 2015 in China. Therefore, as an example of SMPD advantage for scenario
development, two scenarios are here developed under
assumptions that railway construction planning would
have been successfully carried out, increase rate of NPP
would be 0.49 gC m−2 year−1 and elevation on national
level has little change. The two scenarios are distinguished into I and II. In the scenario I, it is supposed
that the urban population proportion would be 57.82%
in 2015, annual increase rate of highway construction
would be 0.25% during the period from 2010 to 2015,
and population would be 1457.84 million persons in
2015 (Fig. 15). In the scenario II, the urban population
proportion would be 46.9%, large-scale highway construction during the period from 2010 to 2015 would
be temporal stagnation and there would be 1417.78
million persons in 2015 (Fig. 16). Both scenarios show
that population might greatly float from the western and
middle regions to the eastern region of China (Table 2).
The more rapid the urbanization and transportation development would be, the bigger the population floating
speed would be.
4. Discussions
The translation of model results into geographical
patterns is already under rapid development by use
of geographical information system (GIS) (Jørgensen,
2002). SMPD is such a method that integrates spatial and non-spatial information from various sources
such as remote sensing, statistics, ecosystem research
network, various monitoring systems, and investigation on-the-spot by means of GIS. In addition to
SMPD, GIS-based methods that have been developed for integrating economic and ecological information in recent years include spatial detailed
Biotope Landscape Model (Muenier et al., 2004),
GIS-extended nitrate pollution model (Matě jı́ček et
al., 2003), FORRUS-S model for forest management
(Chumachenko et al., 2003), Model of Hierarchical
Patch Dynamics (Burnett and Blaschke, 2003), GISbased Erosion Productivity Impact Calculator Model
(Tan and Shibasaki, 2003), Optimisation Methodology for land use patterns (Seppelt and Voinov,
2002), the multi-disciplinary integrated model system
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
consisting of the models ProLand, ELLA and SWAT
(Weber et al., 2001), and Spatial EPIC (Priya and
Shibasaki, 2001). Comparing with other GIS-based
methods, SMPD not only pays attention to the situation of relative elements at the site of generating grid cell itself but also calculates contributions
of other grid cells by searching the surrounding environment of the generating grid cell. For instance,
in the case of this paper, a search radius of 200 km
is defined in the process of generating each grid
cell. Although the Optimisation Methodology adopted
a grid search strategy and performed a grid search
through the entire control space assuming a homogeneous land use and identical fertilizer amounts for
each cell, it aimed at searching temporal change information of each grid cell in a series of maps during the simulation period of 551 days instead of
searching the surrounding environment of the grid
cell.
Human population distribution in China is mainly
determined by geographical location such as distance from cities and transport infrastructures
and environmental conditions such as net primary productivity and elevation. Research results of
wildlife population distributions (Ji and Jeske, 2000;
Krivan, 2003; Westerberg and Wennergren, 2003)
show that distributions of wildlife population are
characterized by geographical location and seasonal variations of movement patterns. Movement
behaviour of wildlife populations is impacted by environmental conditions and land uses. In addition
to predator–prey dynamics and interspecific competition, physical texture, food resources, weather
conditions and landscape structure as well as differences in the magnitude of these factors have a
direct effect on wildlife population distributions.
Therefore, wildlife population distribution is also determined by geographical locations and environmental conditions generally. But some specific factors
such as urbanization and transportation infrastructures, which might negatively contribute to some
wildlife population distributions, have a positive effect on human population distribution; other factors
such as net primary productivity have a similar effect on both human population and wildlife population. It seems that grid generation method employed in
SMPD could be used to simulate wildlife population
distribution.
477
Acknowledgments
This work is supported by National Basic Research
Priorities Program (grant no. 2002CB4125) of Ministry
of Science and Technology of the People’s Republic
of China and by Projects of National Natural Science
Foundation of China (grant no. 40371094).
References
Adams, J., 1968. A Population Map of West Africa, Graduate School
of Geography Discussion Paper No. 26. London School of Economics, London.
Ahuja, D.V., Coons, S.A., 1968. Geometry for construction and display. IBM Syst. J. 7, 188–205.
Burnett, C., Blaschke, T., 2003. A multi-scale segmentation/object
relationship modelling methodology for landscape analysis.
Ecol. Model. 168, 233–249.
Chen, H., Zhang, W.C., 2000. Transport Geography of China. Science Press, Beijing (in Chinese).
China Mayor Association, 2002. National Report of China Urban
Development. Xiyuan Press, Beijing (in Chinese).
Chumachenko, S.I., Korotkov, V.N., Palenova, M.M., Politov, D.V.,
2003. Simulation modelling of long-term stand dynamics at
different scenarios of forest management for coniferous-broadleaved forests. Ecol. Model. 170, 345–361.
Cignoni, P., Montani, C., Rocchini, C., Scopigno, R., 1998. Zeta: a
resolution modeling system. Graphical Models Image Process.
60, 305–329.
Dobson, J.E., Bright, E.A., Coleman, P.R., Durfee, R.C., Worley,
B.A., 2000. LandScan: a global population database for estimating populations at risk. Photogrammetr. Eng. Remote Sens. 66
(7), 849–857.
Foestner, W., 1982. On the geometric precision of digital correlation.
Int. Arch. Photogrammetr. Remote Sens. 24 (3), 176–189.
Gruen, A.W., 1985. Adaptive least squares correlation: a powerful
image matching technique. S. Afr. J. Photogrammetr. Romote
Sens. Cartogr. 14 (3), 175–188.
Heipke, C., 1997. Automation of interior, relative and absolute orientation. ISPRS J. Photogrammetr. Remote Sens. 52, 1–19.
Hu, H.Y., 1983. Discussions on Distribution of Population in China.
Eastern China Normal University Press, Shanghai (in Chinese).
Hu, H.Y., 1935. The distribution of population in China. Acta Geogr.
Sin. 2 (2), 32–74 (in Chinese).
Institute of Geography of Chinese Academy of Sciences, 1987. The
Population Atlas of China. Oxford University Press, Hong Kong.
Institute of Population and Labor Economics of CASS, 2001. Almanac of China’s Population. House of Almanac of China’s Population, Beijing (in Chinese).
Ji, W., Jeske, C., 2000. Spatial modeling of the geographical distribution of wildlife populations: a case study in the lower Mississippi
River region. Ecol. Model. 132, 95–104.
Jiang, Z.H., 1998. China Population Projection. China’s Population
Press, Beijing (in Chinese).
478
T.X. Yue et al. / Ecological Modelling 181 (2005) 461–478
Jørgensen, S.E., 2002. Integration of Ecosystem Theories: a Pattern.
Kluwer Academic Publishers, Dordrecht, pp. 68–71.
Krivan, V., 2003. Idea free distributions when resources undergo population dynamics. Theor. Population Biol. 64, 25–
38.
Lieth, H., Whittaker, R.H., 1975. Primary Productivity of the Biosphere. Springer-Verlag, New York.
Liseikin, V.D., 1999. Grid Generation Methods. Springer-Verlag,
Berlin.
Liu, J.Y., Yue, T.X., Wang, Y.A., Qiu, D.S., Liu, M.L., Deng, X.Z.,
Yang, X.H., Huang, Y.J., 2003a. Digital simulation of population
density in China. Acta Geogr. Sin. 58 (1), 17–24 (in Chinese).
Liu, J.Y., Liu, M.L., Zhuang, D.F., Zhang, Z.X., Deng, X.Z., 2003b.
Study on spatial pattern of land-use change in China during
1995–2000. Sci. Chin. (Ser. D) 46, 373–384.
Liu, M.L., 2003. Dynamic Responses of Terrestrial Ecosystems to
Carbon Cycle Under Impact of Climate Change and Land-use
Change in China. Post-Doctoral thesis of Institute of Geographical Sciences and Natural Resources Research of CAS.
Liu, M.L., 2001. Land-use/Land-cover Change and Vegetation Carbon Pool and Productivity of Terrestrial Ecosystems in China.
Doctoral thesis of Institute of Geographical Sciences and Natural Resources Research of CAS.
Lo, C.P., 2001. Modeling the population of China using DMSP operational linescan system nighttime data. Photogrammetr. Eng.
Remote Sens. 67 (9), 1037–1047.
Matě jı́ček, L., Benešová, L., Tonika, J., 2003. Ecological modelling
of nitrate pollution in small river basins by spreadsheets and GIS.
Ecol. Model. 170, 245–263.
Morrison, D., 1962. Optimal mesh size in the numerical integration
of an ordinary differential equation. J. Assoc. Comput. Machin.
9, 98–103.
Moore, I.D., Grayson, R.B., Ladson, A.R., 1992. Digital terrain modelling: a review of hydrological, geomorphological, and biological application. In: Beven, K.J., Moore, I.D. (Eds.), Terrain Analysis and Distributed Modelling in Hydrology. John Wiley & Sons,
Chichester, pp. 7–34.
Muenier, B., Birr-Pedersen, K., Schou, J.S., 2004. Combined ecological and economic modeling in agricultural land use scenarios.
Ecol. Model. 174, 5–18.
Mustaffar, M., Mitchell, H.L., 2001. Improved area-dased matching by using surface gradients in the pixel co-ordinate transformation. ISPRS J. Photogrammetr. Remote Sens. 56, 42–
52.
National Bureau of Statistics of People’s Republic of China, 2001.
China Statistical Yearbook. China Statistics Press, Beijing.
Priya, S., Shibasaki, R., 2001. National spatial crop yield simulation using GIS-based crop production model. Ecol. Model. 136,
113–129.
Research Center for Population of CASS, 1985. Almanac of China’s
Population. House of Almanac of China’s Population, Beijing
(in Chinese).
Sabin, M., 1990. Commutative surprises in curve and surface theory.
In: Creasy, C.F.M., Craggs, C. (Eds.), Applied Surface Modelling. Ellis Horwood, New York, pp. 9–18.
Sarti, A., Tubaro, S., 2002. Image-based surface modeling: a multiresolution approach. Signal Process. 82, 1215–1232.
Seppelt, R., Voinov, A., 2002. Optimisation methodology for land use
patterns using spatially explicit landscape models. Ecol. Model.
151, 125–142.
Shipley, D.G., 1990. A 3-D finite element mesh generator. In: Creasy,
C.F.M., Craggs, C. (Eds.), Applied Surface Modelling. Ellis Horwood, New York, pp. 19–24.
Stein, M.L., 1999. Interpolation of Spatial Data. Springer-Verlag Inc,
New York.
Sidorov, A.F., 1966. An algorithm for generating optimal numerical
grids. Trudy MIAN USSR 24, 147–151 (in Russian).
Stott, J.P., 1977. Review of surface modelling. In: The Institution of
Civil Engineers Surface (Ed.), Modelling by Computer. Thomas
Telford Ltd for The Institution of Civil Engineers, London, pp.
1–8.
Sutton, P., Elvidge, C., Obremski, T., 2003. Building and evaluating
models to estimate ambient population density. Photogrammetr.
Eng. Remote Sens. 69 (5), 545–553.
Sutton, P., Roberts, D., Elvidge, C., Baugh, K., 2001. Census from
Heaven: an estimate of the global human population using nighttime satellite imagery. Int. J. Remote Sens. 22 (16), 3061–
3076.
Sutton, P., Roberts, D., Elvidge, C., Meij, H., 1997. A comparison of
nighttime satellite imagery and population density for the continental United States. Photogrammetr. Eng. Remote Sens. 63
(11), 1303–1313.
Sutton, P., 1997. Modelling population density with night-time satellite imagery and GIS. Comput. Environ. Urban Syst. 21 (3/4),
227–244.
Tan, G., Shibasaki, R., 2003. Global estimation of crop productivity
and the impacts of global warming by GIS and EPIC integration.
Ecol. Model. 168, 357–370.
Tobler, W., Deichmann, U., Gottsegen, J., Maloy, K., 1997. World
population in a grid of spherical quadrilaterals. Int. J. Population
Geogr. 3, 203–225.
Urban Society and Economy Survey Team of National Bureau of
Statistics of People’s Republic of China, 2001. Urban Statistical
Yearbook of China. China Statistics Press, Beijing (in Chinese).
Weber, A., Fohrer, N., Möller, D., 2001. Long-term land use
changes in a mesoscale watershed due to socio-economic
factors—effects on landscape structures and functions. Ecol.
Model. 140, 125–140.
Westerberg, L., Wennergren, U., 2003. Predicting the spatial distribution of a population in a heterogeneous landscape. Ecol. Model.
166, 53–65.
Year Book House of China Transportation and Communications,
2001. Year Book of China Transportation and Communications.
Year Book House of China Transportation and Communications,
Beijing (in Chinese).
Yue, T.X., Du, Z.P., Liu, J.Y., 2004. High precision surface modelling.
Prog. Nat. Sci. 14 (2), 83–89 (in Chinese).
Yue, T.X., Wang, Y.A., Chen, S.P., Liu, J.Y., Qiu, D.S., Deng, X.Z.,
Liu, M.L., Tian, Y.Z., 2003. Numerical simulation of population
distribution in China. Population Environ. 25 (2), 141–163.
Zhao, S.Q., 1986. Physical Geography of China. John Wiley & Sons,
New York.
Zhang, S.Y., 1997. Population Geography of China. Business Press
House, Beijing (in Chinese).