中文摘要：

摘要：针对现有区域空间格局演化对影响要素及其作用关系研究缺失的问题，本文通过分位数回归建立江苏省2000—2010年县域经济的变量关系模型，并在结合ESDA空间关联模型的基础上对其空间演化格局进行研究，得出以下结论：①变量揭示的区域格局演化的空间关联特征有强化的表现，但不同变量在不同时段分异特征不同。②ESDA探测发现经济发展与变量演化过程可能的潜在关系不同而出现不同类型区空间对应程度的差异，分位数回归给出了不同变量对不同分位地区边际效应影响程度不同的细致信息，从而证实了上述关系的存在。③分位数回归最优拟合值揭示区域经济演化格局与实际吻合度较高，即形成“苏锡”核心热点区和“淮宿”核心冷点区的“圈层结构”，且低值簇内部的演化分异趋势大于高值簇而表现出核心冷点区显著的“内陆化”与次核心冷点区的“沿海化”特征，“高值簇”与“低值簇”分界线一直稳定于苏中地区。

英文摘要：

The present research on the evolution of regional spatial pattern has focused less attention to the interaction relationship between the spatial pattern variable and its influencing factors. This paper aims to establish a variable-relationship model of economic development at county level in Jiangsu from 2000 to 2010, and uses the ESDA analysis model to conduct a study on the evolution process of its spatial pattern. The conclusions can be drawn as follows： （1） The spatial association features of all variables in the regional pattern evolution tend to be more remarkable, but different variables differ from each other in spatial differential features in different periods. （2） The ESDA detects that the variables in economic development and its influencing factors have different spatial corresponding degrees in the evolution process, which indicates different probable potential relationships between them. Furthermore, quantile regression gives different marginal effect degrees of different influencing factors on the economic development during 0.1-0.9 quantile points, which confirms the existence of the relationship mentioned above. （3） The result achieved by quantile regression optimal fitting reveals the evolution pattern of regional economic development is featured by the following aspects. Firstly, a ＂layered structure＂ of the ＂Suzhou-Wuxi＂ core hotspot and the ＂Huaian-Suqian＂ core coldspot will be formed. Secondly, the evolution pattern exhibits a greater evolution differentiation trend in low value cluster than in high value cluster, which manifests .the prominent inland characters of core coldspot areas and the prominent coastal characters of sub-core coldspot areas. Lastly, the boundary of high and low value clusters is generally stable in the central Jiangsu. The results obtained above show that the evolution pattern of regional economic development by quantile regression optimal fitting has a high consistency with the realistic spatial evolution pattern.