关键词: 空间自相关, Morans I值, Getiss G*值, 人口格局

Abstract:

n the evolution of urban, distribution of population in Shanghai has been changing,which is already aroused universal attention and researcher's interest.Based on the population statistic data on the scale of streets in 1990 and 2000,indexes such as Moran's I and Getis's G* were used to analyze spacetime pattern of population distribution of Shanghai,including interoutline road and the scope of urban districts.So,the hotspots,which took on acute characters in the distribution and variation of the population density,were explored from 1990 to 2000.The results show that population centralization is enhanced on the scale of urban districts from 1990 to 2000,high population density is in the urban districts and the low population density in the country.In the center of urban districts,population density is relatively dispersed, which direction is inside of the urban districts.But the discrepancy between urban and rural is increasingly enlarged,which is an urbanization pattern from rural area to urban districts and the process constantly rising.On the scale of the center of urban districts,the density of population is constantly declining.Absolute quantity of population is persistently decreasing in the center of urban districts,while persistently increasing in the periphery of urban districts.So,there are characteristics of obvious suburbanization,and spatialcentrifugal urbanization in the urban area of Shanghai.This paper shows that statistic analysis of spatial autocorrelation should be a good method to explore the hotspots and inner mechanism of distribution,centralization and chang of population.

Key words: spatial autocorrelation, Morans I, Getiss G*, the pattern of population