Tao Shou, Sidan Yao, Qianyu Hong, Jingwen Mao, Yangyang Yuan
As the driving force of China’s green development, cities play a pivotal role in carbon sequestration, with their green and blue spaces jointly influencing both carbon sequestrations and carbon emissions. Yet, most existing studies rely on linear analyses, limiting the capture of nonlinear characteristics and overlooking cross-city differences in spatial configurations. Variations in spatial structures, morphology, and distribution of blue–green spaces may lead to divergent sequestration mechanisms, highlighting the need for comparative research. This study selects five high-density cities in the middle and lower Yangtze River Basin (2000, 2010, 2020) as case studies. Using the XGBoost-SHAP model, we investigate the correlations between blue–green space patterns and carbon sequestration benefits across cities. Results show that key indicators vary by city: patch shape complexity, patch area, and connectivity significantly affect sequestration benefits across all cases, while patch proximity, size, shape, and spatial aggregation matter in specific cities. This study provides a reference for optimizing urban blue–green space configurations from the perspective of carbon sequestration benefits and offers a direction for further exploration of their underlying mechanisms. At the planning level, the study identifies key indicators influencing carbon sequestration across different urban forms, providing a scientific basis for context-specific optimization of blue–green space structures and for promoting low-carbon and resilient urban development.