This study examines how spatial context influences the relationship between environmental performance (lower carbon emission intensity) and financial performance (higher profitability) in real estate investment trusts (REITs). Analyzing 375 REITs across 21 economies during 2017-2023, we identify a negative, non-linear relationship between carbon emission intensity and profitability. However, this relationship is significantly weaker for REITs headquarted in major urban agglomerations. We attribute this to three mechanisms: reduced environmental differentiation in dense markets, elevated operational costs, and systemic environmental externalities. In contrast, REITs outside agglomerations show a stronger negative relationship, driven largely by Scope 2 emissions. Our findings reveal the importance of geographical context in shaping green financial outcomes and highlight the need for spatially sensitive policy design. We also discuss how AI can enhance emission monitoring, hotspot identification, and resource optimization, strengthening the environmental-financial link in future real estate management.