Forest or grassland? A quantitative analysis of urban residents' green exposure preference by using multi-temporal mobile signal data

Geng, H.; Lin, T.; Bodegom, P.M. van; Hu, M.; Zhen, Y.; Jia, Z.; ... ; Lin, J. (2025)

Article / Letter to editor

Urban forests and grasslands provide diverse services from their unique characteristics. Optimizing green spaces by understanding urban residents’ preferences is a critical challenge for sustainable city development amid limited land resources. However, the mechanism influencing exposure across various types of green remains unclear. This study utilized multi-temporal mobile signal data from Shanghai to quantify the exposure intensity (EI) and density (ED) for forests and grasslands. These metrics addressed the gap by revealing spatiotemporal variations in exposure preference (EP) and related socioeconomic influences. Specifically, the study addressed two key questions: (1) Do urban residents exhibit preferences between forests and grasslands in terms of EI and ED? (2) How do socioeconomic features influence these preferences? Results showed: (1) Forests had almost double the annual EI (542.86 p/h) and ED (2.69 p/m²/h) of grasslands (P < 0.001). However, grasslands in central regions exhibited significantly higher ED (13.60 vs. 11.83 p/m²/h; P < 0.001). (2) Commercial House (34.4 % importance) and Sports & Recreation (15.7 %) maximized green exposure, while Road Furniture reduced it. (3) Evening exposure peaks in central regions extended by 1 hour due to commercial-cultural synergies. Forest ED, highly driven by Commercial House, clustered in central cores and specific non-central communities, whereas Road Furniture most negatively impacted central periphery communities. These findings directly inform differentiated urban planning strategies: forests should prioritize improving accessibility to sustain prolonged exposures, while grasslands need spatial optimization to accommodate peak social demand. By aligning green space planning with socioeconomic drivers, cities can enhance the effectiveness of their service under land constraints.

• Urban green space
• Green exposure
• Exposure preference
• Machine learning
• Multi-Scale Geographically Weighted Regression

• All authors: Geng, H.; Lin, T.; Bodegom, P.M. van; Hu, M.; Zhen, Y.; Jia, Z.; Zhang, J.; Guo, X.; Chen, Y.; Lin, M.; Cai, J.; Lin, J.
• Date: 2025-06-01
• Journal: Urban Forestry and Urban Greening
• Volume: 108
• DOI: doi:10.1016/j.ufug.2025.128826
• Link: https://www.sciencedirect.com/science/article/pii/S1618866725001608?via%...
• Sponsorship: NWO
• Grant number: 482.19.704