文献类型：期刊文献

中文题名：城乡地域系统“三生”空间识别与特征——以北京市海淀区为例

英文题名：Spatial identification and feature analysis of "production-living-ecological" spaces in urban-rural regional system: A case study of Haidian district of Beijing

作者：刘小茜[1];马思克[1,2,3];杨磊[4];杜宜霖[1]

第一作者：刘小茜

机构：[1]北京联合大学应用文理学院,北京100191;[2]云南大学国际河流与生态安全研究院,昆明650500;[3]云南大学生态与环境学院,昆明650500;[4]中国科学院生态环境研究中心区域与城市生态安全全国重点实验室,北京100085

第一机构：北京联合大学应用文理学院

年份：2025

卷号：80

期号：10

起止页码：2720-2736

中文期刊名：地理学报

外文期刊名：Acta Geographica Sinica

收录：;北大核心:【北大核心2023】；

基金：国家自然科学基金项目(42271112,42230718);国家留学基金项目(202308110126)。

语种：中文

中文关键词：“三生”空间;城乡梯度;城乡地域系统;地理大数据;空间识别;北京市海淀区

外文关键词："production-living-ecological"spaces;urban-rural gradient;urban-rural regional system;multi-source geographical big data;spatial identification;Haidian district of Beijing

摘要：“三生”空间的精准识别是城市功能区优化、宜居城市建设和城乡协调发展的基础。当前“三生”空间功能识别的研究中,对融合空间和交互空间的辨识及其定量刻画不足,且缺乏对不同城市化程度城乡地域系统的区别化认知,亟需在区分城乡地域系统的基础上融合多源地理大数据,构建体系完整、量化精准的空间功能识别方案。本文以北京市海淀区的路网小区为基本判读单元,基于POI、土地覆盖、手机通讯等数据构建了空间功能强度指数(SFS)、空间功能覆盖指数(SFC)和空间功能交互指数(SFI),并使用决策树设定功能判别优先级,结合景观格局指数揭示了城乡“三生”空间的多尺度(全局、类别和斑块)分异特征。通过多源地理大数据的融合,充分利用不同数据的优势实现数据支撑内容的互补,从而实现对各基本判读单元功能属性的精准识别。研究发现:(1)研究区全局尺度上存在城乡梯度差异。随城市化程度的升高呈现出空间类型多样性增加、同类空间聚集度降低、形状复杂度增加、地块密度增大等空间规律。(2)类别尺度上空间功能结构差异显著。单一的生态空间和生活空间分别在乡村地域系统和城市地域系统中集中分布,单一生产空间在不同地域系统下的形态特征因主导产业不同而存在差异,融合空间的类型和形态特征存在典型城乡差异,交互空间具有“职—住”分离特征且广泛分布在城乡过渡带。(3)斑块尺度的景观指数,在更精细的角度辨析了“三生”空间各类别斑块的形态和功能组合特征,有效识别出具有最高空间多样性和最小空间聚集性的特殊城市功能区。本文通过创新数据融合方式,有效发挥了多源地理大数据的优势,实现了对复杂地域系统的“三生”空间功能判读,为国土空间规划和城市化研究提供了科学支持。
Accurately identifying the "production-living-ecological" spaces (PLES) is the premise of urban functional zone optimization, livable cities construction, and urban-rural development balancing. Existing research on identifying spatial distribution of PLES is insufficient in identifying and quantitatively characterizing integration space and interactive space. There is an urgent need to establish methods that consist of complete system and can accurately quantify spatial functions by integrating multi-source geographic data based on existing rural-urban regional systems. Thus, this present constructed three indices, including Spatial Function Strength index (SFS), Spatial Function Coverage index (SFC), and Spatial Function Interaction index (SFI). These indices were calculated using the Point of Interest (POI) data, land cover, and cell phone communication data using the road network community as the fundamental unit. The PLES in Haidian district, Beijing was identified by determining the priority of the three indices with the decision tree interpretation method. In addition, this study compared the spatial distribution and characteristics of functions at the global, class, and patch levels using landscape indices. The results include: (1) The functional zones at different regional systems were significantly different. Along the urban-rural gradient, the diversity spatial types and complexity of shapes, and the patch density increased, but the spatial aggregation in the same class decreased. (2) Individual ecological and living spaces showed concentrated distribution within the rural and urban regional systems, while the spatial characteristics of individual production space in different regional systems were quite different in terms of dominant industry type. Within the integrated spaces, the type and characteristics were different between rural and urban areas. For interactive spaces, they were mainly distributed in the transitioning spaces between rural and urban zones, characterized by "separation between living and working spaces". (3) At the patch level, landscape indices complemented and supported the spatial pattern for the system and classes of the PLES at a finer scale. In addition, landscape indices effectively identified special city functional regions with the largest spatial diversity and smallest spatial aggregation. The result of this study realized system interpretation and spatial identification of multi-functionality for PLES at the urban-rural regional systems based on innovated methods for integrating data in illustrating the advantages of multi-source geographic big data, which provided supports in terms of data and methods for territorial spatial planning and management.