Assessing Climate Stress, Biodiversity and Urban Quality of Life in Île-de-France Using Artificial Intelligence and GIS

Urban areas are increasingly exposed to climate stress, particularly heat extremes, while simultaneously facing biodiversity loss and growing inequalities in quality of life. Understanding the spatial interactions among these dimensions is essential for designing effective and equitable urban adaptation strategies. This study develops an integrated AI–GIS analytical framework to assess climate stress, urban biodiversity, and quality of life across the Île-de-France metropolitan region. Using a combination of satellite-derived land surface temperature, extreme heat indicators, vegetation indices, landscape fragmentation metrics, and socio-environmental variables, the analysis captures spatial heterogeneity at the departmental level. Geographic information systems and spatial statistics are employed to identify heat stress hotspots and ecological connectivity patterns, while machine learning models (Random Forest, XGBoost, and artificial neural networks) are used to model nonlinear relationships and assess predictor importance. The results reveal pronounced intra-metropolitan disparities in climate stress, with densely urbanized departments experiencing higher land surface temperatures, more frequent heatwave exposure, and greater population vulnerability. Areas with lower vegetation cover and higher ecological fragmentation exhibit amplified thermal stress, whereas well-connected green infrastructures contribute to climate mitigation and improved quality-of-life outcomes. Artificial intelligence models confirm the critical role of vegetation indices, built-up density, and green space accessibility as key determinants of urban heat intensity. Integrated clustering further identifies distinct spatial typologies combining climate stress, biodiversity conditions, and quality-of-life indicators. Overall, the findings highlight the necessity of integrated urban planning approaches that simultaneously address climate adaptation, biodiversity conservation, and environmental equity. The proposed framework offers a transferable decision-support tool for climate-resilient metropolitan planning under ongoing climate change. JEL Classification · Q54 – Climate; Natural Disasters; Global Warming · Q57 – Ecological Economics; Ecosystem Services · R14 – Land Use Patterns; Urban Spatial Structure · R58 – Regional Development Policy · C45 – Neural Networks and Related Topics