Spatial patterns and interactions among multiple cultural ecosystem services across urban greenspaces

Urban greenspaces (UGS) deliver substantial benefits to human wellbeing by providing valuable ecosystem services. Prior research on UGS has been primarily focused on provisioning and regulating services, with comparatively fewer studies explicitly addressing cultural ecosystem services (CES), presumably due to conceptual and methodological challenges in their characterization and quantification. Social media data have emerged as novel datasets that could provide new insights into the quantification of these intangible, highly context-specific, but critically important CES. In this study, we merged multiple platforms, including TripAdvisor and Google Maps that are among the most comprehensive user-generated datasets, to map and quantify the spatial distribution of 11 CES. Employing named-entity recognition models, this study extracted 60,156 textual entities related to CES from scraped reviews, allowing us to categorize 30,599 reviews into different CES types across 426 urban greenspaces. Our research demonstrated substantial spatial heterogeneity in the presence and diversity of CES and identified six key CES bundles, revealing more occurrences of CES synergies than tradeoffs across UGS. Geographical random forest models were applied to determine the relative importance of natural landscape elements, biodiversity proxies, and human utility metrics in explaining the spatial heterogeneity of CES. We found that factors such as greenspace size, tree cover percentage, biodiversity, and water features emerged as strong predictors of CES provision. Our study provides a roadmap and research framework for understanding and quantifying CES in urban settings and has implications for the sustainable planning and management of UGS to improve social wellbeing through the contribution of diverse CES.