区域发展战略对国土空间格局演化的影响——以云南省为例

地区国土空间格局演化不仅受自然资源禀赋等先天因素的影响,还受到集聚经济等后天因素的影响。区域发展战略通过改变当地的基础设施条件、交通通达性、功能布局等经济活动空间集聚的条件,重塑地区经济地理。资源环境承载力和国土开发适宜性评价（“双评价”）作为优化国土空间格局的重要基础和参照,过于强调资源本底,仅仅评估现状,也缺乏区域发展战略和重大工程项目对资源环境承载能力和国土空间开发适宜性产生的影响和变化的评估。因此,单纯依靠“双评价”对未来国土空间演化趋势和土地供需展开研判不够准确,对集聚经济的重视亦不足。本文首先构建一个空间经济学的分析框架,探讨和辨识国土空间演化过程的多种影响因素和机制,建立一个基于空间经济学理论的国土空间演化模拟分析平台,并通过人工神经网络（ANN）和元胞自动机（CA）整合实现机器学习和演化模拟。其次,在对比分析与校正的基础上,设置5种未来区域发展战略情景,对云南省全域进行高分辨率模拟分析,并据此对“双评价”结果进行修正。研究表明：①包含集聚经济因素的模拟结果更能反映空间经济活动的规律,且有更高的模拟精度。②各种区域战略的实施均有利于优化国土空间的开发格局,其中综合统筹和农业优先两种情景更有利于地区充分利用比较优势、因地制宜开展国土空间的开发工作。③基于综合统筹情景修正的“双评价”结果,进一步识别出了云南省国土空间适宜建设区、储备建设区、生态保护极重要区和不适宜建设区,在补充现有“双评价”结果的同时进一步对云南省现有国土空间格局进行优化。本文阐释了国土空间格局演化的空间经济机理,为新时代国土空间格局的优化以及“双评价”的修正,提供科学支撑和政策建议。

Impact of regional development strategies on territorial spatial evolution: A case study of Yunnan Province

The evolution of regional spatial patterns is influenced not only by First Nature such as natural resources endowment, but also by Second Nature such as agglomeration economy. Regional development strategies reshape regional economic geography by changing local infrastructure conditions, transportation accessibility, functional layout, and other conditions of spatial agglomeration of economic activities. The “Double Evaluation” is an important basis and reference for delineating the “three zones and three lines” and optimizing the land use spatial patterns. However, the Double Evaluation emphasizes the natural resources conditions and only assesses the current situation. It also lacks the assessment of the impact and changes of regional development strategies and engineering projects on the carrying capacity of the environment and the suitability of territorial spatial development. Therefore, the predications of future spatial evolution trend and land supply and demand are inaccurate if we rely solely on the Double Evaluation. This study first constructed an analytical framework of spatial economics to explore and identify various influencing factors and mechanisms of the spatial evolution. Then, a spatial evolution simulation and analysis platform based on spatial economics theory, artificial neural network (ANN), and cellular automata (CA) were established to realize the evolution simulation for Yunnan Province. Based on the comparative analysis and modification, this study set up five scenarios of future regional development strategies and conducted a high-resolution simulation analysis for the province, and revised the Double Evaluation result accordingly. The results of our study show that: (1) The simulation results with the agglomeration economy factor can better reflect the pattern of spatial economic activities and have higher simulation accuracy. (2) The implementation of various regional development strategies is conducive to optimizing the territorial spatial development patterns; and the two scenarios of comprehensive integration and agricultural priority are more favorable for cities to make full use of their comparative advantages and carry out spatial development according to local conditions. (3) Based on the comprehensive and overall development scenario, the revised Double Evaluation result further identifies suitable construction area, reserve construction area, extremely important ecological protection area, and unsuitable construction area, optimizing the existing territorial spatial development patterns of Yunnan Province.