摘要
分析了传统的人口空间分布密度衰减函数———指数型和Gauss型 ,指出了其应用的局限性 ;对于有两个中心以上的城市 。
This paper points out the essentiality of spatial distribution from statistical population data after analyzing the usual methods for describing urban population density.A framework is proposed for spatial distribution from statistical population data.1)The study region is divided into fine grids; 2)Identify a centroid for each zone and assign the population data to this centroid; 3)Interpolate a grid surface by using the population data of this set of points and by a interpolation techniques. This paper uses kernel estimation to break down count data published for one fixed set of irregular areal units to a fine grid which can then be re_aggregated to other sets of areas and thus used in conjunction with information on an alternative areal base. Finally,an experiment on a city in China's Mainland shows that kernel estimation is feasible and effective for modeling spatial continuous surface of population density for multicentric cities.
出处
《武汉大学学报（信息科学版）》
EI
CSCD
北大核心
2002年第3期301-305,共5页
Geomatics and Information Science of Wuhan University
基金
国家计委 2 0 0 0年高技术应用资助项目
关键词
人口
空间分布化
GIS
population
spatial distribution
GIS
