CAJ | 학술논문

空间自相关分析是一种空间统计的方法。可以揭示区域变量的空间结构形态。通过对2000-2003年期间上海市乡镇人口密度变化率的空间相关性分析，发现上海市近几年人口分布空间变动格局表现为以中心城区为中心的环状分布模式，中心城区和远郊区人口密度变化小。人口增长缓慢，而近郊区人口密度变化大，人口增长迅速。全市人口密度变化率存在着显著的空间集聚现象。四种类型的空间关联在空间分布上具有明显的规律性。研究结果表明，目前上海市人口郊区化趋势明显。高人口密度区域在不断向外扩张，人口分布与经济、交通等社会发展因素的关系日益密切。这些研究结果说明利用空间自相关统计方法能够扩展和加强对人口分布空间模式的研究．从而为经济总体布局和城市化建设提供辅助参考作用。
공간자상관분석시일충공간통계적방법。가이게시구역변량적공간결구형태。통과대2000-2003년기간상해시향진인구밀도변화솔적공간상관성분석，발현상해시근궤년인구분포공간변동격국표현위이중심성구위중심적배상분포모식，중심성구화원교구인구밀도변화소。인구증장완만，이근교구인구밀도변화대，인구증장신속。전시인구밀도변화솔존재착현저적공간집취현상。사충류형적공간관련재공간분포상구유명현적규률성。연구결과표명，목전상해시인구교구화추세명현。고인구밀도구역재불단향외확장，인구분포여경제、교통등사회발전인소적관계일익밀절。저사연구결과설명이용공간자상관통계방법능구확전화가강대인구분포공간모식적연구．종이위경제총체포국화성시화건설제공보조삼고작용。
Spatial autocorrelation statistic is a kind of spatial statistics, which can explain the spatial structure of regional variables. In particular, rate of change of population density is a key index to describe alternation of population distribution. This article firstly summarized the theory of spatial autocorrelation, especially reviewed some important concepts, such as spatial weighted matrix, Moran＇s I, local Moran＇s I and so on. Then it analyzed the situation of population distribution in shanghai from 2000 to 2003. It is diificult to discover the relationship of regions by using traditional natural break statistical method through describing the general characteristics of population density. While using Spatial autocorrelation methods to analysis population density of villages and towns, the results show that, spatial population distribution of the city is characterized by the increasing ring of population surround central zone of Shanghai ; there are significant phenomena of spatial clustering in the study area; local version of Moran＇s I is particularly useful for decomposing the pattern of spatial association into four categories, which have obvious regularity in terms of spatial distribution. The result also indicates that population suburbanization of Shanghai develops very quickly and a lot of near suburbs will become highly - density regions. Population distribution has become increasingly related with economic, traffic and other social development factors. Also, this research indicates that, Spatial autocorrelation analysis can be extended and strengthened by spatial statistical analysis, allowing for the deepen exploration, analysis, manipulation and interpretation of spatial patterns and spatial association of population distribution. In summary, it can provide reference for overall economic planning and urbanized construction.