Seasonal variations and spatial mechanisms of 2D and 3D green indices in the central urban area

Green indices are critical indicators of urban greening quality, essential for optimizing and managing urban green spaces. However, most existing research focuses on vegetation growth during the growing season, limiting its ability to capture the seasonal dynamics and spatial impact factors of urban greening. To address this gap, this study uses the central urban area of Beijing as a case study. First, it extracts the 3D green indices from street view imagery and integrates them with the 2D green indices derived from remote sensing data to assess seasonal variations. The spatial clustering of the green indices and the impact of socio-economic and biophysical factors are then analyzed. Finally, multiscale geographically weighted regression is applied to examine the spatial heterogeneity of environmental impact factors on green indices distribution. The results show that the composite green index was 0.359 in summer and 0.304 in winter, based on subdistrict averages. Across subdistricts, the index declined by 13.27% on average from summer to winter. The green indices exhibited more substantial spatial autocorrelation in winter, with 2D green indices showing higher clustering than 3D indices. Socio-economic factors were the primary drivers of green indices overall, while biophysical factors had a more substantial impact during summer. Accordingly, the Central Business District and DingFuzhuang clusters should be prioritized for planting evergreen species and implementing a multi-layered vegetation structure. This study provides a quantitative foundation and practical insights for seasonal planning and adaptive optimization of urban green space system in densely developed urban environments.