|
|
| |
|
|
| 本文已被：浏览 51次 下载 0次 |
|
|
|
| 轨道交通滨海站点周边建成环境与城市活力影响机制探索 ——以大连市为例 |
|
陈飞, 胡秦兰, 韩名洋
|
|
大连理工大学
|
|
| 摘要: |
| 轨道交通滨海站点周边建成环境对塑造多维城市活力具有复杂且非线性的影响；本文聚焦滨海城市，针对该影响机制展开深入研究。以大连滨海区域13个典型轨道站点1km缓冲区为研究对象，借助“5D模型”，融合滨海城市特点，构建了包含建筑密度、功能混合度、公交接驳度、海岸线距离等在内的10项建成环境指标，并以社会活力、经济活力与文化活力为因变量，形成三维测度体系。基于多源时空大数据，耦合LightGBM机器学习算法与SHAP可解释性框架，系统地揭示建成环境对城市活力的非线性阈值效应与交互机制。研究发现：(1)关键建成环境变量，如到地铁站距离、到市中心距离、到海距离、功能密度、滨海路径密度等对多维活力的影响存在显著阈值效应，表现为积极型、消极型或多阶段波动等多种非线性形态；（2）滨海特征在城市活力的塑造中具有独特的作用机制：当到海距离较小时，因土地利用受限与功能配置不足，地段的经济活力会遭到抑制；而随着到海距离增大，会逐渐形成“滨海最优带”，兼顾生态保护与功能集聚；另外，滨海路径密度的提升可以显著缓解远海距离对文化活力的不利影响。(3)关键变量间存在强交互效应，如市中心距离与海岸距离协同促进经济活力，功能密度与地铁可达性共同增强社会活力。研究成果揭示了滨海站点活力生成的复杂空间机制，为空间优化设计提供量化依据，对推动滨海城市人本化更新与可持续发展具有重要理论与实践意义。 |
| 关键词: 轨道交通滨海站点 建成环境 城市活力 LightGBM 非线性效应 |
| DOI： |
| 分类号:TU984 |
| 基金项目:辽宁省哲学社会科学规划基金项目“人居环境高质量发展视域下辽宁南部地区公众亲海空间评价与提升研究” |
|
| Exploring the Influence Mechanism of the Built Environment Around Coastal Rail Transit Stations on Urban Vitality —Taking Dalian as an example |
|
CHEN Fei, Hu Qinlan, Han mingyang
|
|
Dalian University of Technology
|
| Abstract: |
| As vital spatial conduits for land-sea interaction, coastal cities are undergoing a transformation of their shoreline functions—shifting from traditional industrial transportation to multifunctional “living waterfronts.” In this process, public recreation, commercial services, and cultural tourism are increasingly clustering in waterfront areas, placing higher demands on the planning and design of coastal spaces. Rail transit stations, serving as pivotal nodes connecting urban hinterlands with coastal frontiers, have evolved far beyond basic commuting functions. They increasingly act as core engines for revitalizing coastal areas, enhancing tourism experiences, and shaping urban identities. The quality of built environment configurations around stations directly impacts the sustainability of coastal vitality and comprehensive competitiveness. However, existing research predominantly focuses on the relationship between urban vitality and built environments in inland cities and general settings. Systematic and in-depth exploration remains lacking regarding the complex, nonlinear mechanisms linking the built environment around stations to multidimensional urban vitality within the unique coastal context. Therefore, revealing the nonlinear patterns and spatial mechanisms through which the built environment surrounding coastal rail transit stations influences urban vitality holds significant theoretical value and practical urgency.Taking the typical coastal city of Dalian as a case study, this research selected 13 representative transit stations within a 1,000-meter radius of the shoreline along the residential waterfront in Dalian's main urban area, specifically within the section from Qixianling to Jinshitan. These stations represent the closest points to the shoreline while serving areas with high population density, effectively embodying typical scenarios of residential waterfront spaces. Unlike stations serving solely commuting functions, these sites feature integrated residential, recreational, and tourism functions with excellent rail accessibility.Theoretical Framework and Measurement System Building upon relevant theoretical foundations, this study constructs an urban vitality measurement system encompassing three dimensions: social vitality, economic vitality, and cultural vitality. Social vitality is characterized by intraday population density data collected via Baidu Huayan population big data, reflecting travel and gathering intensity. Economic vitality is depicted through active merchant review density extracted from Dianping.com, measuring commercial activity and consumption potential. Cultural vitality is measured by quantifying cultural and entertainment facilities from AutoNavi POI data, capturing the spatial distribution of coastal public cultural resources. For built environment measurement, the study drew upon the classic “5D” model—density, diversity, design, destination accessibility, and proximity to public transit—while integrating coastal spatial characteristics. This culminated in a 10-core indicator system featuring metrics like “distance to coastline” and “coastal path density” that highlight coastal scenario traits.In terms of methodology, to overcome the limitations of traditional linear models in capturing complex nonlinear relationships and interaction effects, this study coupled advanced machine learning algorithms with an explainable modeling framework. Specifically, for model comparison, linear regression, random forest, and XGBoost were selected as controls. Using R2 and RMSE as primary metrics, the LightGBM (Light Gradient Boosting Machine) algorithm model was ultimately chosen for its superior performance. The SHAP (SHapley Additive exPlanations) interpretability framework was then applied to analyze the trained LightGBM model. SHAP values quantify each feature variable's contribution to model predictions, while its dependency plots clearly illustrate nonlinear relationships and threshold points between individual features and the target variable. SHAP interaction values effectively reveal synergistic or antagonistic interactions among different feature variables.Through SHAP dependency graphs and interaction effect analysis, the study found:(1) Variables across different built environment dimensions exhibit varying relative importance in influencing the three vitality dimensions, though distance factors dominate in coastal scenarios. For economic and cultural vitality, “distance to city center” is the most influential variable; for social vitality, “functional density” exerts the strongest impact. Additionally, “distance to subway station” and “distance to coastline” demonstrate significant importance across all vitality dimensions. In contrast, design-related indicators like vegetation coverage and floor area ratio exhibit relatively limited influence. This highlights that in generating vitality within coastal site areas, locational accessibility—particularly relative proximity to urban cores, transportation hubs, and coastal resources—is a key spatial determinant.(2) Nonlinear relationships and distinct threshold effects generally exist between the built environment and multidimensional vitality. In the economic vitality dimension, “distance to the coastline” exhibits a pronounced spatial threshold effect: within 0 to 365 meters, overly proximate coastal locations may inhibit economic vitality due to potential land-use restrictions or functional homogeneity; Between 365 and 1486 meters, the impact shifts to a significant positive effect, forming an “optimal coastal zone” that balances ecological sensitivity and functional clustering. Beyond 1486 meters, the positive effects of coastal location gradually diminish. The impact of “distance to subway stations” on economic vitality follows an inverted U-shaped curve, with the optimal range around 400-500 meters. Beyond 758 meters, the spillover effect turns negative. Regarding social vitality, “functional density” exhibits an activation threshold of approximately 280 units/km2; exceeding this value significantly boosts social vitality. “Vegetation coverage” exhibits an optimal threshold of approximately 0.49, with excess coverage potentially inhibiting vitality. Regarding cultural vitality, “distance to coastline” also has an optimal range of approximately 507–1873 meters, where proximity initially promotes but eventually inhibits vitality. These specific threshold points provide precise quantitative guidance for refined spatial design.(3) The study uncovered strong interactive effects among key built environment variables, indicating that vitality emerges from the complex coupling of multiple factors rather than the isolated influence of any single element. Key interactions include: For economic vitality, “distance to city center” and “distance to coastline” exhibit synergistic effects—sites simultaneously possessing central and coastal locations demonstrate significantly higher economic vitality. For social vitality, “functional density” and “distance to subway stations” exhibit a complementary moderating relationship, meaning that areas with high functional density significantly enhance social vitality when conveniently connected to subway systems. For cultural vitality, “coastal path density” effectively moderates the impact of “distance to coastline,” as a well-developed coastal slow-travel network can mitigate the adverse effects of greater distance from the coast on cultural vitality.The significance of this study lies in theoretically transcending traditional linear paradigms. It systematically reveals, for the first time, the complex nonlinear, threshold-based, and interactive mechanisms linking built environments to multidimensional vitality within the specific context of coastal rail transit stations. This deepens our understanding of the intricate human-land interactions in coastal urban spaces. Practically, the identified thresholds—such as subway station economic influence zones, optimal coastal belts, and functional density activation thresholds—along with their interactive patterns, provide precise, science-based decision support for coastal urban spatial planning, station-adjacent urban design, land-use functional layouts, slow-travel system development, and policy formulation. Finally, the study highlights current limitations in the depth of social interaction data and the measurement of intangible dimensions of cultural vitality, pointing to areas for future research advancement. |
| Key words: Coastal Rail Transit Stations Built Environment Urban Vitality LightGBM Nonlinear Effects |
|
|
|
|
