1SPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS”, Bangkok, May 23-25, 2001 
72 
SPATIAL DEVELOPMENT RESEARCH OF LARGE CITY BASED ON GIS SPATIAL ANALYSIS 
Anrong DANG Qizhi MAO Xiaodong WANG 
Center for Science of Human Settlements, Tsinghua University 
Qinghua Yuan, Beijing 100084, P. R. China 
Tel: 86-10-62783328 / Fax: 86-10-62770314 
E-mail: Danronq® tsinqhua.eclu.cn 
KEYWORDS: Spatial Development, Large City, GIS, Spatial Analysis, Spatial Interaction 
ABSTRACT 
According to the theory of spatial interaction and model of comprehensive attraction between cities, the spatial development of Beijing, 
the Capital of China, was studied in this paper supported by GIS spatial analysis functions. There are four parts in the article. At first, the 
theory basis and model characteristic was reviewed and discussed. Two important models were quoted to calculate the current attraction 
between cities and development potentiality of Beijing. Then, the method and workflow were put forward which include four steps, such as 
database preparing, spatial analyzing, potentiality calculating, and comprehensive analyzing. Based on the method mentioned above, the 
calculation process was described one by one, such as determining development axes of Beijing, defining model parameters of attraction, 
selecting mass of cities around Beijing, measuring distance between Beijing and other cities, and calculating development potentiality of 
Beijing. Finally, comprehensive analysis was done and some results were put forward. The conclusion is that the future spatial 
development direction of Beijing is southeast axis which facing many cities, such as Tianjin, Langfang, and Huanghua. 
Beijing, as the Capital of China, its spatial development in the Xueqiang, etal, 1997). 
21th century is interested by many persons in China and 
overseas (References 2-6). According to the theory of spatial After the calculation of comprehensive attraction between cities 
interaction and model of comprehensive attraction between cities, using model (3), the development potentiality of each city can be 
the authors have analyzed the spatial development of Beijing summarized by model (4). 
supported by GIS spatial analysis functions. 
cities. In fact, this is not always true. So, some comprehensive researching and the possibility of data acquiring, Beijing region 
models were put forward in practice, such as “comprehensive databases were setup supported by GIS and DBMS. The 
1. THEORATICAL BASIS 
(4) 
City, especially capital, places an important role in social In model (4): L represents the development potentiality of city i. 
development. Spatial interaction between cities is important for Dm is the distance of city i itself which is measured by average 
city development. The earlier scientist who interested in the radius of city i or the half of the distance between city i and its 
spatial interaction of cities is W. J. Reilly. He worked out a famous nearest neighbor. The means of other parameters are just same 
model which named “retail attraction rules” based on the law of as in model (3). 
universal gravitation (Refers with Eq. (1), Reilly, 1931). 
(1) 
2. METHODOLOGY FRAMEWORK 
According to the principle of spatial interaction between cities, the 
demand of spatial development research for Beijing, and the 
natural environment and social economy situation of the region, a 
framework of research method is worked out as shown in Fig. 1 
based on the characteristics of GIS. The framework includes four 
steps: data preparing, spatial analyzing, productivity calculating, 
and comprehensive analyzing. 
Based on the model of W. J. Reilly, P. D. Converse worked out 
another model named “Break Point Model” gravitation (Refers 
with Eq. (2), Converse, 1949). 
2.1 Data Preparing 
Both model (1) and model (2) regard population as the mass of "^ e ^ rs *. s * e P ' s *° co ^ e p\ ranted data. On the necessity of 
attraction model” (Refers with Eq. (3)). 
WP'XWjPj) 
D h ij 
(3) 
databases include spatial database (such as administrative 
divisions, topographic contour, transportation system, residential 
distribution, and so on) and attribute database (such as 
population statistical data, staff and workers, gross domestic 
products, and so on). These data are the basis for further 
calculating and analyzing. 
2.2 Spatial Analyzing 
In model (3): L represents the quantity of spatial interaction 
between city i and city j. Wj and Wj refer to mass weight factor of On the basis of residential distribution and transportation system, 
city i and city j. Pj and Pj refer to mass of city i and city j. D^ seven possible development axes and their radiation regions 
represents the distance between city i and city j. The order of have been determined. Then buffer analysis has been done alone 
distance, b, is named friction coefficient. Among the parameters the transportation system, and overlay analysis and topographic 
in model (3), mass of city can be population or other factors analysis have been done in each axis radiation region. Based on 
related to application problem. Distance between cities can be the spatial analysis above, the amount of residential and their 
transportation distance or cost, communication distance or cost, spatial distribution in each region have been obtained, along with 
or time distance. Mass weight and friction coefficient must be the land slope composition and their spatial distribution in each 
determined according to real problem. Case studies revealed that region. Furthermore, take the relationship of residential 
friction coefficient is always changed from 0.3 to 0.5 (Xu distribution and transportation system into account, the shortest