Optimizing urban park ecological benefits: Linking 3D structure to function with LiDAR and GAMs

Urban green spaces are important providers of ecosystem services, yet systematic quantification of how different elements of stand structure contribute to distinct ecological benefits remains limited. We used the China Green Expo Garden in Zhengzhou, which contains approximately 108,241 trees representing 121 species, as a case study and integrated field surveys with multi-source data. We quantified the influence of three structural groups and twelve structural variables on ecological benefits, including cooling, particulate removal, carbon sequestration, oxygen production, and rainfall interception - using Spearman correlation analysis, Random Forest models, and Generalized additive models (GAMs). The results indicate that the park's mean annual total ecological benefit (TEB) is approximately CNY 19,262.6 per hectare. The relative contributions of horizontal, vertical, and species composition structural components to TEB were about 64.5%, 25.3%, and 10.2%, respectively. Canopy cover exhibited a threshold effect: the optimal canopy cover for maximizing overall ecological benefits was about 60%–70%, whereas the canopy cover optimal for carbon sequestration was lower, at 40%–50%. High-density plantings of smaller trees were more effective at enhancing ecological benefits than strategies focused solely on increasing individual tree diameter. Spatial-structure optimization based on GAMs could increase each of benefits by 10%–30% under moderate optimization intensity and by 25%–45% under high-intensity optimization. This study provides quantitative evidence and practical framework for prioritizing ecosystem-service objectives and guiding evidence-based green-space design at the park scale.