The use of spatial autocorrelation to analyze changes in spatial distribution patterns of population density in Jiangsu province, China

Spatial autocorrelation is an important spatial statistical method in Geographic Information Systems, which can reveal spatial structures and patterns of regional variables. With this method, this article calculates the population density´s spatial correlation model for county-level units in Jiangsu province of China in 2001, 2002, 2003 and 2004, explores and demonstrates every county-level´s global and local spatial correlation, and analyzes factors influencing the patterns of Jiangsu province´s population distribution. The study results show that the population density´s global Moran´s I reaches significant spatial positive correlation, and presents apparent spatial agglomeration. Furthermore, the population density´s local Moran´s I cluster maps display that local spatial agglomeration of the population distribution in Jiangsu is also very significant. The study results demonstrate that statistical analysis of spatial autocorrelation can well reveal population distribution patterns, the concentration of population and its hot changes and internal mechanism, and it´s important for government macro-decision making such as rational distribution programs for population and urbanization policy. According to this conclusion, this article proposes some suggestions to promote the sustainable development of future population and economy in Jiangsu.