文章编号： 1672-5328（2024）06-0007-10

李雯茜1, 2，卞长志2，许奇3，董天一4，赵一新2
(1. 北京交通大学交通运输学院，北京100044；2. 中国城市规划设计研究院，北京100037；3. 北京交通大学综合 交通运输大数据应用技术交通运输行业重点实验室，北京100044；4. 北京交通大学自动化与智能学院，北京 100044)

摘要： 城市轨道交通带来的土地价值提升归因于交通邻近溢价、可达性提升溢价以及车站周边的建 设溢价三方面，后两项对土地价值提升的影响研究尚不充分。融合多源城市大数据提出城市轨道交 通TOD建成环境的“5D”变量及其计算方法，放松既有研究中房价与其影响因子的线性关系假 设，采用基于机器学习模型XGBoost的特征价格非线性模型和局部解释方法SHAP，分析TOD建成 环境对土地价值提升的非线性影响及其空间分布特征。针对北京市的案例研究表明：TOD建成环境 变量对土地价值的影响程度达64.30%，其中公共交通可达性相对重要度达17.66%，较车站邻近性 更重要；TOD建成环境对土地价值的影响存在明显的非线性关系和阈值效应。TOD建成环境的空 间分布特征表明针对不同区域的车站用地开发应采取因地制宜的差异化发展策略，城市外围区域车 站应注重资源聚集，中心城区车站应侧重出行品质和环境质量提升。

关键词： 城市轨道交通；建成环境；土地价值提升；可达性；机器学习；阈值效应；北京市

中图分类号： U491

文献标识码：A

The Nonlinear Impact and Spatial Characteristics of TOD Built Environment on Land Value: A Case Study of Beijing

LIWenxi1, 2, BIAN Changzhi2, XU Qi3, DONG Tianyi4, ZHAO Yixin2
(1. School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China; 2. China Academy of Urban Planning & Design, Beijing 100037, China; 3. Key Laboratory of Transport Industry of Big Data Application Technologies for Comprehensive Transport, Beijing Jiaotong University, Beijing 100044, China; 4. School of Automation and Intelligence, Beijing Jiaotong University, Beijing 100044, China)

Abstract： The improvement of land value brought by urban rail transit can be attributed to three aspects: the premium of adjacent transit stations, the premium of accessibility improvement, and the construction premium around the station. However, the research on the impact of the latter two factors on the improvement of lane value remains insufficient. In response, integrating multi-source urban big data, this paper proposes the“5D”indicator system and its calculation methodology for the Transit-Oriented Development (TOD) built environment of urban rail transit. Relaxing the assumption of linear relationship between house price and its influencing factors in the exiting research, the paper employs a nonlinear model for feature pricing based on the XGBoost machine learning algorithm and the local interpretation method SHAP to analyze the nonlinear impact of TOD built environment on the promotion of land value and its spatial distribution characteristics. Case studies in Beijing demonstrate that the TOD built environment around rail stations significantly influence the land value, accounting for 64.30%, and the relative importance of transit accessibility holds is 17.66%, which is more important than station proximity. The impact of TOD built environment on land value has obvious nonlinear relationship and threshold effect. The spatial distribution characteristics of TOD built environment indicate that station land development strategies should be tailored to specific regions, according to local conditions, the stations in the peripheral areas of cities should pay attention to the accumulation of resource, and the stations in the central urban area should focus on the improvement of travel quality and environmental quality.

Keywords： urban rail transit; built environment; land value uplift; accessibility; machine learning; threshold effect; Beijing