ISPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS”, Bangkok, May 23-25, 2001 
74 
Fig.3. Spatial Development Axes of Beijing 
3.3 Model Parameters Defining 
When you apply Model (3) and model (4), two kinds of model 
parameters must be defined before calculating attraction 
between cities. One is weight factor of city mass, and another is 
friction coefficient of city distance. Because ten index were 
selected to represent the mass of cities, the weight factor of city 
mass is very difficult to define relayed on city mass directly. 
However, the administrative class of city has non-substituted 
affection to the spatial interactive between cities. Talking about 
the friction coefficient of city distance, it is based on the 
transportation type and class. So, the weight factor of city mass 
was defined by administrative class of cities, while the friction 
coefficient of city distance was defined by the transportation type 
and class. The detail values of the parameters were shown in 
table 2. 
Table 2. Parameters of City Attraction 
Weight Factor of City Mass 
Friction Coefficient of Transportation 
Administrative 
Class 
Weight Factor 
T ransportation 
Type & Class 
Friction 
Coefficient 
Capital 
1.0 
Electric Railway 
0.6 
Municipality 
0.9 
Double-line Railway 
0.7 
Prefecture 
0.8 
Single-line Railway 
0.8 
City 
0.7 
Super Highway 
0.9 
County 
0.6 
Main Highway 
1.0 
3.4 City Distance Measuring 
The simplest distance between cities is straight-line length, but it 
is nearly meaningless for interaction analysis between cities. The 
real distance between cities is the shortest transportation mileage 
that could mirror the spatial interaction of cities. On the basis of 
transportation system analyzed in background section, the 
distance between Beijing and other cities were measured by 
means of GIS network analysis function. During the 
measurement of distance between cities, the transportation type 
and class were recorded at the same time. So, one to one 
relationship between the distance and friction coefficient was 
established for further calculation. 
3.5 Mass Data Processing 
Ten social and economic factors were selected to represent 
mass of cities, which include urban population, staff and worker, 
average wages, budgetary revenue, budgetary expenditure, 
gross domestic product, gross output value of industry, gross 
output value of agriculture, and so on. Because of the different 
unit and value, the calculation result will be not true. In order to 
remove the affection of unit and value, data standardization has 
been processed by two kinds of ways: percentage and average. 
Table 3. Data Standardization of City Mass 
Original Data of City Mass 
Standardized Data of City Mass 
Total Population 
Urban Population 
Staff and Workers 
Average Annual Wage 
Gross Domestic Products 
GOV of Industry 
GOV of Agriculture 
Budgetary Revenue 
Budgetary Expenditure 
Retail Sales of 
Consumer Goods 
Percentage of Total population 
Percentage of Urban population 
Percentage of Staff and Workers 
Average Annual Wage per Capita 
Per Capita GDP 
Per Capita GOV of Industry 
Per Capita GOV of Agriculture 
Percentage of Budgetary Revenue 
Percentage of Budgetary Expenditure 
Percentage of Per Capita 
Retail Sales of Consumer Goods 
3.6 Potentiality Calculating 
After all of above preparing process, the current attraction 
between Beijing and other cities for each kind of city mass have 
been calculated by means of model (3) and integrated spatial 
database. Then, comprehensive attraction and development 
potentiality of each seven direction were calculated and put in 
order (refer to table 4). The comprehensive attraction means that 
how much Beijing will be attracted by each city in each axis 
radiation region. And the development potentiality means that 
how much potentiality Beijing will be developed in the direction. 
Table 4. Comprehensive Attraction and Potentiality 
Development 
Axes 
Comprehensive 
Attraction 
Sequence 
order 
Development 
Potentiality 
Sequence 
order 
Southeast 
3863.46 
1 
62194.52 
1 
Southwest 
2429.42 
4 
60994.62 
2 
East 
2795.02 
2 
58824.76 
3 
Northwest 
1968.91 
5 
47259.27 
4 
South 
1938.85 
6 
46838.47 
5 
Northeast 
2670.33 
3 
46397.79 
6 
West 
1009.52 
7 
23732.61 
7 
4. DEVELOPMENT ANALYSIS 
According to the calculating value of comprehensive attraction, 
development potentiality and their sequence order, development 
analysis has been done and some results can be obtained as 
follows. 
(1 )From the view of development potentiality value, the sequence 
order is that southeast axis is the first, while the southwest axis is 
the second, and the east axis is the third. It means that the future 
development direction of Beijing must be inclined to southeast. In 
that axis, the current city system and transportation network is 
much better than others. At the same time, Tianjin harbor and 
Huanghua Harbor are two important bases for international trade. 
In the second development direction, cities are more 
concentrated than others, while the transportation network, the 
social economic situation, and the industry system are just well. 
And in the third development axis, both New Tangshan and 
Qinghuangdao harbor have much attraction to Beijing. 
(2)From the view of comprehensive attraction value, the 
sequence order is that southeast axis is the first, while the east 
axis is the second, and the northeast axis is the third. This is not 
totally the same to development potentiality. Although both the 
attraction and the potentiality are good in east direction, the 
southwest only has more potentiality but attraction because of the 
concentrated cities. At the same time, the southeast has both the 
maximum development potentiality and comprehensive 
attraction. 
Obviously, the final conclusion is that the future spatial 
development direction of Beijing is southeast axis which facing 
Tianjin, Langfang, and Huanghua cities.