The Game Between Quality Induction and Traffic Constraint: A Non-Linear Threshold Study of Park Travel Carbon Emissions from an Urban–Rural Differentiation Perspective

As global decarbonization strategies pivot towards the burgeoning sector of recreational mobility, the tension between the attractive force of high-quality amenities and the constraining capacity of transport infrastructure in urban parks has emerged as a critical planning dilemma for high-density metropolises. To disentangle this game mechanism, this study proposes a integrated Dual-Diagnostic Framework that synthesizes a modified gravity model, Grouped OLS regression, and an explainable XGBoost-SHAP algorithm to identify non-linear thresholds under spatial heterogeneity. Leveraging empirical data from Tianjin, a representative high-density metropolis, the analysis reveals a distinct bimodal distribution of carbon emissions from travel to comprehensive parks, confirming a fundamental structural divergence between urban and suburban mobility patterns. Crucially, the non-linear diagnosis uncovers a dominant Facility Configuration Induction mechanism within the suburban interface; here, park scale acts as the primary driver of excess travel, with its induction effect often overriding the mitigation potential of public transit until a specific critical mass is achieved. Consequently, the results identify a rigid threshold for bus station density alongside optimal intervals for park scale, providing quantitative benchmarks and differentiated governance strategies to resolve the paradox between park quality and carbon intensity.