地理研究

地理研究 ›› 2004, Vol. 23 ›› Issue (1): 97-103.doi: 10.11821/yj2004010012

• 论文 • 上一篇    下一篇

区域人口密度函数与增长模式:兼论城市吸引范围划分的GIS方法

王法辉1, 金凤君2, 曾光2   

  1. 1. 美国北伊利诺大学地理系;
    2. 中国科学院地理科学与资源研究所 北京100101
  • 收稿日期:2003-06-18 修回日期:2003-09-22 出版日期:2004-02-15 发布日期:2004-02-15
  • 作者简介:金凤君,(1961-),内蒙古人,研究员。从事经济地理研究工作。E-mail:jinfj@igsnrr.ac.cn.
  • 基金资助:

    中国科学院知识工程创新项目(KZCX23071)

Analyzing regional density functions and growth patterns in China with a GIS-based method delineating influential regions of cities

WANG Fa-hui1, JIN Feng-jun2, ZENG Guang2   

  1. 1. Department of Geography, Northern Illinois University, DeKalb, USA;
    2. Institute of Geographic Sciences and Natural Resources Research,CAS,Beijing 100101,China
  • Received:2003-06-18 Revised:2003-09-22 Online:2004-02-15 Published:2004-02-15

PDF (PC)

赞

被引次数

219 | 19

摘要：

区域人口密度函数是研究中心城市对周围区域影响的有效方法。为了控制自然因素对人口密度分布的影响 ,本文的研究地域范围限定在我国四个主要的平原地区 :东北、华北、两湖平原和四川盆地。城市吸引范围是基于重力模型用GIS方法划分的。利用 1982年至 1990年的人口资料进行的模型模拟的结果表明 ,我国区域人口分布特征同西方国家一样 ,呈现出距离衰减特征 ,即随离城市的距离的增加 ,人口密度逐渐下降。由于城市经济增长的原因 ,城市人口增长快于边远地区 ,呈现出向心集聚的趋势。但不同区域集聚程度不一 ,核心城市人口快速增长的区域 ,腹地的人口增长比较迟缓 ;而核心城市人口增长速度一般的区域 ,腹地近乎是同步增长。

关键词: 区域密度函数, 区域增长模式, 城市吸引范围划分, GIS, 中国

Abstract:

Taking advantage of a GIS data set of county-level administrative regions and the National Population Census data in 1982 and 1990, this research analyzes the regional growth patterns in China through the change of regional density functions. To minimize the influence of physical environments on population densities, the study areas are limited to four major plains of China: the Northeast China, North China,Hubei-Hunan Plains and the Sichuan Basin. These plain areas are defined approximately according to cultivation ratios at the county level. A gravity-based model is used to delineate the influential regions of 17 cities. In other words, the influence of a city on a county is positively proportional to the city's population size but negatively proportional to the distance between them. A county is included in the influencial region of a city if this city exerts the largest influence on the county among all surrounding cities. The model is implemented in a GIS environment. In China, regional densities decline with distance from a city, similar to western countries. Four simple bivariate functions are tested: (1) linear, (2) exponential, (3) reverse exponential and (4) power functions. Among the four functions, function (3) or D r =a+blnr fits the regional density patterns in China the best. This is different from urban density patterns, which are best captured by the (negative) exponential function. Based on the change of fitted density curves over time, regional growth patterns can be identified. The results show that in all 17 regions, areas close to central cities grew faster than remote areas, described as a trend of centralization. However, regions with strong core growth are generally associated with stagnant hinterlands; and regions with moderate core growth are usually matched by similar growth rates in the hinterlands. This indicates that most regions in China are still on the process of centralization, i.e., fast growth in core areas (urban and suburban) at the expense of peripheral areas. This trend is strongest in regions where the central cities have gained the fastest growth.

Key words: regional density functions, regional growth patterns, delineation of influential regions, GIS, China