Multi-Perspective Evaluation of Urban Green Views: Spatial and Street-View Data Integration in Sudirman Central Business District, Indonesia

Abstract

Urban green spaces (UGSs) are critical for enhancing urban livability and sustainability by providing both ecological and human-centered benefits. This study integrates spatial landscape metrics and the street-level visibility of greenery (measured through the green view index [GVI]) in order to evaluate the structural and visual characteristics of UGSs in a dynamic urban area – specifically, the Sudirman Central Business District (SCBD) of Jakarta, Indonesia. The analysis focuses on examining the roles of landscape metrics such as area, perimeter, compactness, shape index, and elongation in influencing the GVI and its spatial variability across different types of urban green spaces (including parks, green corridors, and open spaces). The results indicated that larger and more compact UGSs significantly contributed to higher GVI levels (thus, reflecting better visual greenery), while elongated and fragmented green spaces exhibited greater variability and lower visibility. Non-linear relationships (assessed through random forest regression and SHAP analysis) further revealed the complex interactions between GVI and landscape metrics, thus emphasizing the importance of incorporating advanced statistical approaches. The limitations that are related to data quality, temporal coverage, and spatial heterogeneity are also discussed, thus highlighting opportunities for future research for addressing these challenges through multi-temporal analyses and spatially explicit models. By bridging the gap between the spatial configurations and visual perception of UGSs, this study contributes to sustainable urban-planning strategies that are aimed at optimizing green spaces for ecological functionality and human well-being.