利用广义加性模型解析长江中游城市群城镇用地扩展驱动力

研究城镇用地扩展的时空驱动机制对城市群的科学可续发展与国土空间规划具有重要意义.利用广义加性模型(General Additive Model,GAM),选取社会经济因子、自然因子和邻域因子,解析长江中游城市群1990～2015年城镇用地扩展驱动力的时空分异特征,挖掘各子城市群的核心驱动力并对其进行功能识别.结果表明:(1)25年间,长江中游城市群驱动因子时空差异显著.核心城市武汉、长沙、南昌极化效应明显,涓滴效应逐渐显现,但城市群仍未打破行政区划壁垒.(2)三大子城市群的核心驱动因子差异显著.其中,人口和地均GDP对各子城市群城镇用地扩展影响最大,武汉城市圈和环鄱阳湖城市群的空间溢出效应高于长株潭城市群,且城镇用地扩展模式渐由单核驱动发展为双核驱动.(3)核心驱动因子与城市群城镇用地扩展之间具有显著的非线性关系,且呈现地域性和阶段性特征.(4)GAM模型能够解释城镇用地扩展过程中各因子之间的非线性关系,较Logistic和Nonliner-Logistic模型具有更高的拟合优度和验证精度,表明了其探索非线性驱动力的优势.

Analysis of Driving Forces of Urban Land Expansion in Urban Agglomeration in the Middle Reaches of the Yangtze River Based on General Additive

Studying the spatial-temporal driving mechanism of urban land expansion is of great significance to the scientific sustainable development and national spatial planning of urban agglomerations. Based on the General Additive Model (GAM), this paper selects socio-economic factors, natural factors and neighborhood factors to analyze the spatial-temporal differentiation pattern of the urban land expansion driving forces in urban agglomeration in the middle reaches of the Yangtze River from 1990 to 2015. We then explore the core driving force of each sub-urban agglomeration and identify its function. The results show that: (1) In the past 25 years, the spatial and temporal differences of driving factors of urban agglomeration in the middle reaches of the Yangtze river were significant. The core cities Wuhan, Changsha and Nanchang had obvious polarization effects and gradually appeared as the trickle-down effect. However, the urban agglomeration still had not broken the administrative division barriers. (2) The core driving factors of the three sub-urban agglomerations manifested significantly different. Among them, the population and GDP per area had the greatest impact on the urban land expansion in each sub-urban agglomeration. The spatial spillover effect of the Wuhan metropolitan area and the urban agglomeration around Poyang Lake was higher than that of the Changsha-Zhuzhou-Xiangtan urban agglomeration, and the expansion of urban land use changed from a single-core development mode to dual-core development mode. (3) There was a significant non-linear relationship between the core driving factors and the urban land expansion of urban agglomerations, and the characteristics of regionalism and stage were presented. (4) The GAM model can explain the nonlinear relationship between various factors in the process of urban land expansion, which had higher goodness of fitting and verification accuracy than Logistic and Nonliner-Logistic models, indicating its advantages in exploring non-linear driving forces.