关键词: 马尔科夫链, 俱乐部趋同, 绿洲经济, 新疆

Abstract: Based on per capita GDP data of 87 cities and counties of Xinjiang in 1949—2007, this paper analyzes the spatio-temporal evolution of economical pattern by means of spatial probability model and ArcGIS software platform. According to per capita GDP data and the economic development level Xinjiang is divided into 4 classes of developmental types: upper level, middle-upper level, middle-lower and lower level. The temporal transfer trends and the spatial transfer pattern were analyzed with Markov Chain Transition Matrix and ArcGIS, respectively. The dynamic mechanism of the spatio-temporal evolution of economics development level was also analyzed based on spatial Markov chain. The results are as follows. (1) The "convergence club" phenomenon of economic development level commonly happened in Xinjiang in 1949—2007, especially for the upper level and the lower level of economic development. The main driving force was the transformation of economic structure from Primary Industry to Secondary Industry and Tertiary Industry. (2) The transfer of economic development types showed different directions in various stages and different density between varied regions; obvious transfer in regions between adjacent levels and unobvious transfer astride levels. (3) During the Planned Economy Period, economic development background promoted the transformation of economic levels, but played a promoting role in maintaining economic levels during the Market Economy Period. (4) For regions of lower economic development level, the transfer probability between economic levels was small and the negative "Matthew Effect" was significant. Whereas, for regions of upper economic development level, the transfer probability was relatively large and the fragility and inferiority of economic development was significant.

Key words: Markov matrix, club convergence, oasis economy, Xinjiang