Accurately measuring the carbon sink of urban green trees is a prerequisite for improving the carbon sink capacity of green spaces. This article takes Pinghu City as an example, using a single tree survey combined with the green space type method to measure the carbon sink of urban trees, clustering analysis of the carbon sink capacity of tree species, and spatial autocorrelation and heat map analysis of carbon density and carbon sink rate spatial distribution characteristics to assist in the rational layout of urban green spaces. The results showed that: (1) The carbon density and carbon sink rate of subsidiary green spaces were the highest in Pinghu, and the average tree carbon density of urban tree green spaces (22.72 t C/hm²) was at the upper middle level. The average tree carbon sink rate (8.99 t CO₂e a^-1 hm^-2) was higher than that of non-urban forests. (2) The carbon sequestration capacity of Hardwood Ⅱ ((31.13±15.11) kg CO₂e/a) and Hardwood Ⅰ ((30.43±17.56) kg CO₂e/a) was stronger than that of Pinus, Softwood, Cunninghamia, and Cupressus. (3) The carbon density and carbon sink density of trees in urban green spaces exhibited spatial clustering, with high values often distributed in strips along roads or in clusters along water bodies. The methods and conclusions of this study have important theoretical and practical significance for clarifying the carbon background of urban green space trees, managing and enhancing the natural carbon sink function of cities, and promoting urban carbon neutrality.