Analysis On Spatial Correlation And Spatial-temporal Comparison Based On Urban Comprehensive Competitiveness

There are a lot of factors which affect city competitiveness. Thus, city competitiveness show spatial inequality and has obvious characteristics of location, which makes classical statistics encounter puzzles. Many of the fundamental theories of spatial statistics are developed from classical statistics, but its spatial models and analytical methods are significantly different from classical statistics. In spatial statistics, the valuable observations generally have their own locations. However, with the change of geographical position, the relationship and structure among variables will also change. The article explores and analyzes the spatial structure of city competitiveness in China and the pattern of space-time, which has practical significance for government reasonably planning cities or regions and provides reference for regional economics and other relevant research in urban studies.In this paper, based on studying the urban competitiveness system model Wang Guixin put forward, on the whole country with combined environment characteristic, economic traits, and social characteristic, especially, Consumer Price Index in the last decade, is used to analyze the impact of consumer price change on actual expenditure for living cost of cities.In this way, the urban competitiveness in different years can be compared. Then the conclusions from the model for evaluation of urban competitive are drawn. Secondly, from anal sizing with spatial variation, the characteristic of spatial structure for urban competitiveness is showed, imitating urban spatio-temporal trend surface by comprehensive competitiveness is done. For simulating spatial trend surface, Kriging interpolation has been done. Polynomial trend surface imitating algorithm with geographical factors will improve the precision. The purpose of models fitting and optimizing has been realized. Based on it, spatio-temporal pattern, evolution trend, spatial diffusive and attractive mode of cities in different regions has been analyzed. And finally draw the following conclusions:(1)By computing of the index of spatial autocorrelation, the urban comprehensive competitiveness is positively correlated in distributing spatially, presenting obvious character of spatial clustering pattern directivity. The local spatial autocorrelation analysis reveals that, the cities show the character with high-high concentration in comprehensive competitiveness mainly take coverage to BoHai Sea around, ChangJiang Delta and ZhuJiang Delta, while the cities with "low-low" concentration, mainly locate in the southern of Gansu Province, Ningxia Province, the northeastern of Sichuan Province, and Heilongjiang Province. (2) According to the analysis by spatial variability, urban competitiveness showed obvious anisotropy spatially, and as the change process is bigger, the null decreases year by year, which shows that two different cities gradually connect, as sphere of urban Impact on regions expands increasingly.(3) From the results of analysis with Kriging interpolation, there are three main centers for distributing urban comprehensive competitiveness, namely the region between Beijing and Tianjin, the region following Shanghai, Wuxi and Suzhou, and the whole neighborhoods among Shenzhen, Zhuhai and Dongguan. The overall spatial distribution pattern of urban comprehensive competitiveness extends outward from the three centers. Central plains cities, urban agglomeration in Wuhan and Hu-Bao-e urban groups form the three minor centers.