基于手机信令数据的城内迁居活跃度识别方法研究——以武汉市为例

中国城市的快速发展加速了城市内部人口的居住迁移,精细测度居民迁居行为及其空间分异特征,对于从人的行为视角来分析城市居住空间结构演进过程具有现实意义。本文以武汉都市发展区为例,基于手机信令大数据,提出净活跃度指标和总活跃度指标来量化描述迁居活跃度,并依据两个指标的分类及组合,将居民居住地划分为高迁入型活跃区、平稳型活跃区、高迁出型活跃区、高迁入型非活跃区、平稳型非活跃区与高迁出型非活跃区六种空间类型,进而探讨居民迁居的空间分布特征。结果表明：① 武汉都市发展区内部人口迁移总体呈现出从主城区向近郊区逐步迁移的趋势,且主城区人口总活跃度相较更高。② 主城区以高迁出型活跃区和高迁入型活跃区为主,近郊区则以平稳型非活跃区为主。③ 不同空间类型内的居住类型存在差异：高迁入型活跃区内以新建小区、高校住区、学区房、城中村为主;平稳型活跃区以农村居住地、园区周边住宅为主;高迁出型活跃区以老旧小区、园区周边住宅、城中村、农村居住地为主;非活跃区则以农村居住地为主。本文提出了一种基于时序手机信令大数据的居民迁居活跃度评价指标体系,并以实证研究证明其对于居民迁居地空间类型划分的有效性,研究结果可为相关规划决策部门掌控城市内部的人口迁移特征提供数据支撑、为城市不同区域针对性的进行公共资源配置提供参考依据。

Research on the identification method of relocation activity degree in inner city based on mobile phone signaling data: A case study of Wuhan

The rapid development of Chinese cities has accelerated the residential mobility of inner city. Precisely measuring residents' migration behavior and its spatial differentiation characteristics is of practical significance to analyze the evolution process of urban residential spatial structure from the perspective of human behavior. Taking Wuhan Metropolitan Area as an example, based on mobile signaling big data, this paper proposes the net activity index and the total activity index to quantitatively describe relocation activity degree, and classifies the residential places into six spatial types based on the classification and combination of the two indicators: high immigration and active area, stable and active area, high emigration and active area, high immigration and inactive area, stable and inactive area, and high emigration and inactive area, and then explores the spatial distribution characteristics of residents' migration. The results show that: (1) The population migration within Wuhan Metropolitan Area generally shows a trend of migrating from the main city to inner suburbs, and the total activity degree in the main city is higher than that in the suburbs. (2) The main city is dominated by high emigration and active areas, and high immigration and active areas, while the inner suburbs are dominated by stable and inactive areas. (3) There are differences in residential types among different spatial types: high immigration and active areas are mainly new residential areas, college residential areas, school district housing, and urban villages; stable and active areas mainly consist of rural residential areas and residential buildings around industrial parks; high emigration and active areas are mainly old residential areas, residential buildings around industrial parks, urban villages and rural residential areas; inactive areas are mainly rural residential areas. This paper proposes an evaluation index system of residents' relocation activity degree based on time-series mobile signaling big data, and proves its validity for the spatial classification of residents' migration places by empirical research. The results can provide data support for relevant planning decision-making departments to regulate population migration within a city, and provide reference for public resources allocation in different areas of the city.