<p>Urban green space system serves as a carrier for various economic activities, connecting social and economic elements of urban and rural areas. This role is crucial for achieving carbon-peaking and carbon-neutrality goals. However, clarifying mesoscale green space carbon sequestration and its driving mechanisms remains challenging. Taking Tianjin, China, as a case study, green spaces within the urban growth boundary were identified through supervised classification of GF-1 images and manual digitization based on online imagery. The extracted green spaces include all types, such as park green spaces, affiliated green spaces, square green spaces, protective green spaces, regional green spaces, and other informal green spaces. Based on a carbon density model, carbon stock efficiency was estimated. Principal component analysis and geographically weighted regression were employed to analyze the driving factors. We propose the measures to improve the carbon sequestration of green space and the improvement strategy. The main results are as follows: (1) The distribution of urban growth boundary green space in Tianjin is concentrated in the center and scattered in the periphery. (2) Total carbon sequestration of green space in Tianjin urban growth boundary is 5.68 × 10⁵MgC and the total carbon sequestration density is 2.46 MgC/ha, and that in core of city is larger than the outer edge. (3) The permanent resident numbers, population aged 15–64, and construction land coverage most strongly affect carbon sequestration. A transdisciplinary framework can analyze urban growth boundary carbon sequestration across planning-construction-maintenance phases. This approach could establish theoretical foundations for urban–rural planning while improving regional carbon sequestration balance.</p> Graphical abstract <p></p>