<p>Across modern history, global population dynamics have undergone substantial transformations with significant implications for sustainability, achieving clarity of the dominant demographic factors could hence provide insights for more targeted green transition strategies. Based on the STIRPAT model, and transnational data from 1971 to 2023 covering 196 countries, this research adopts multiple methodologies including Shapley value, FEVD, and ML to decompose complex demographic changes into three categories: scale, structure, and distribution, and examined their impact on global GHG emission. The results suggest that: (1) modern demographic changes are associated with global GHG emission increasement; (2) population spatial density shows greater correlation with GHG than other demographic factors; (3) robust SEM mediation models suggest that urban construction and renewal, pollution transfer, public service provision and digitalization mediate this effect differently; (4) heterogeneity exists in terms of aging phases, affluence and gender equality. By clarifying the relative importance of different demographic dimensions, this study shifts the focus from aggregate population size toward spatial distribution and structural composition. The findings suggest that: amending rural and suburban areas could benefits green transition, correcting excessive regional imbalance is meaningful for sustainable development, a stronger North-South coordinating mechanism including environment transfer payment and green technology sharing is helpful in achieving the SDG 11.</p>