关键词: 变差函数, 空间自相关, 时空格局, 县域经济发展, 中国

Abstract: Studying the spatio-temporal pattern of county economic development is extremely important for revealing the evolutionary mechanism of regional economy and achieves the sustainable development of China's county economy. Although considerable research attention has examined the regional economic pattern of China at the provincial and national levels in a certain time, we currently know the least about the profiles of county economic development across different periods. There is general agreement in the regional geography community that understanding the dynamic pattern of county economic development can provide a scientific base for the policy making and the implementation of regional development planning. Thus, the spatio-temporal pattern of county economic development in China merit further investigation. Based on the 2352 counties' per capita GDP in 1982, 1990, 2000 and 2010, the spatial autocorrelation analysis and variogram were used to investigate the evolutionary characteristics of spatiotemporal patterns of county economic development and to explore the possible mechanism behind the changes in the spatial pattern. Results showed that the county economic development and growth exhibit a positive spatial autocorrelation, which indicates that some counties with similar economic development levels clustered. But the spatial autocorrelation of the economic growth over the past three decades was not obvious. Since 1982, the spatial concentration of the county economy increased gradually and the spatial dependence enhanced over time. The difference of county economic development in eastern area of China is greater than that in the middle and western regions. In all four studied years, there are nine county units with a "high-high" pattern of economic development, i.e., Wujin, Kunshan, Dantu, Taicang, Changshu and Wuxi. Meanwhile, there are about forty-three county units maintaining their locations of "low-low" pattern of economic development for all studied years. Furthermore investigations revealed that the proportion of the county unites with homogeneity economic level increased from 19.56% in 1982 to 27.26% in 2010, whereas unites with heterogeneity level decreased from 3.06% in 1982 to 2.55% in 2010. This result demonstrated that there was an obvious heterogeneity for China's county economic development and the polarization effect in county economic development was enhancing over time. Overall, the hotspot areas of China's economic development mainly clustered in the eastern and northern regions of China, while the hotspot ones were concentrated in its central, southern and southwestern areas. Furthermore, the continuity and self-organization of Chinese county patterns was enhancing, whereas the random components of spatial disparity patterns was decreasing over time, which means that the structural differentiation caused by spatial autocorrelation was becoming more apparent. There was a relatively good homogeneousness in the economic development in the direction of northeast-northwest. The possible drivers behind the county economic development patterns could be partly attributed to its economic location, regional development policy, special resources and background of historical development. Among these possible causes, the regional development strategies may be the external contributors to the evolution in regional economic patterns. These findings have important theoretical and practical significance on narrowing the gap between urban and rural development.

Key words: China, county economic development, spatial autocorrelation, spatial-temporal pattern, variogram