Spatiotemporal patterns and governance effects of China’s urban MSW-induced air pollution in the context of population urbanization: Spatial evidence and machine learning insights

Rapid population urbanization in China has intensified municipal solid waste (MSW) generation and reshaped the spatial distribution of pollution risks, particularly affecting the hydroscape. If not properly managed, this can lead to groundwater contamination and the influx of plastic waste through stormwater runoff into rivers and lakes, disrupting the structure and function of aquatic ecosystems and hindering progress toward Sustainable Development Goals (SDGs). Despite growing scholarly attention, most existing studies are city-specific or factor-isolated, lacking a comprehensive national-level analysis of spatiotemporal patterns and interaction effects. This study used provincial-level datasets from mainland China (2000–2022) to construct a comprehensive air pollution index based on the relative risk coefficient. Standard deviation ellipse and spatial autocorrelation methods are employed to capture spatiotemporal patterns, followed by an XGBoost model integrated with the SHAP method to quantify the interactive response of demographic and governance drivers. The results reveal that the gravity center of air pollution shifted from central provinces toward the southeastern coastal urban belt, with notable clustering in the southeast, while the annual average followed an S-shaped curve. Spatial clustering and polarization intensified significantly over time. Urbanization-related factors exhibited threshold effects, whereby exceeding critical levels leads to pollution reduction or stabilization. Consequently, governance-related factors demonstrate stage-specific effectiveness, and emphasizing the importance of process optimization. Among the interactive drivers, R&D personnel show the strongest interaction dominance, displaying significant coupling synergistic effect with population density and composting capacity, as well as optimal coupling intervals effect with urbanization rate and environmental expenditure. This study addresses a critical research gap by integrating spatial analysis and interpretable machine learning to uncover nonlinear and synergistic mechanisms behind MSW-induced air pollution in rapidly urbanizing contexts. Based on these findings, the study proposes targeted policy strategies, including regional integration of environmental governance, performance-based budgeting, and a shift from disposal capacity expansion to efficiency- and emission-oriented management. These findings provide evidence-based insights for advancing Sustainable Development Goal 13 (Climate Action), which could enhance the resilience and sustainability of environmental systems, aligning with the integrated agenda of nature, technology, and governance.