Integrating Earth Observation and Graph Theory to Evaluate Urban Green Spaces Connectivity Across European Capitals

Résumé fourni par la source

Abstract Context As global urbanization intensifies, Urban Green Spaces (UGS) are pivotal for biodiversity conservation and climate change mitigation. However, comparative assessments of UGS spatial configuration and connectivity across diverse urban landscapes remain limited. Objectives This study aims to assess the spatial arrangement and connectivity of UGS across 28 European capital cities. Additionally, we evaluate how Network Science metrics derived from Graph Theory can complement traditional landscape ecology metrics to provide a more comprehensive understanding of UGS at a large scale. Methods We developed a European Urban Vegetation Map using Earth observation data to classify UGS at 10m resolution across the selected capitals. We then analyzed UGS connectivity for each city utilizing 40 traditional landscape metrics and a Graph-Theory-based approach. Results While traditional landscape metrics effectively quantified fragmentation, they often remain strongly correlated with total vegetation abundance. In contrast, Network Science metrics provided specific insights into UGS functional connectivity, distinguishing the quality of ecological links beyond spatial proximity. This integration allowed us to cluster European capitals into three distinct typologies: unconnected compact cities, large metropolises with complex peri-urban dynamics, and high-connectivity cities with robust networks. These findings demonstrate that graph-based indices effectively complement traditional metrics, highlighting that relying solely on green space percentage is insufficient for assessing the ecological resilience of urban environments. Conclusions These results underscore the relevance of Earth observation-based UGS assessment and demonstrate that graph-based landscape connectivity analysis outperforms simple abundance metrics. Therefore, effective assessment requires integrating structural metrics with graph-based connectivity to support resilient urban biodiversity.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.