Multi-scale determinants of net primary productivity in China's urban agglomerations.

Urban agglomerations are key engines of socio-economic growth and impose profound pressures on ecosystem functioning. However, the relative contributions of different drivers to carbon sequestration dynamics within urban areas during urban agglomeration development remain poorly quantified. Here, we investigated the spatiotemporal dynamics of net primary productivity (NPP) across five major urban agglomerations in China from 2000 to 2020 using the process-based Integrated Biosphere Simulator (IBIS) model and scenario-based simulations. The relative contributions of key drivers, including CO2 fertilization, climate change, and land use/land cover change (LUCC) were then disentangled at both administrative boundary and urban boundary scales. Results showed that CO2 fertilization consistently enhanced NPP, partially offsetting climate-induced declines, whereas LUCC exerted region-specific effects promoting NPP gains in reforested areas but driving substantial loss in rapidly urbanizing regions such as the Yangtze River Delta. At the administrative boundary scale, environmental drivers (CO2 fertilization and climate change) dominated NPP trends, whereas at the urban boundary scale anthropogenic LUCC emerged as the dominant driver of gains (relative contribution >70%). Urban vegetation enhancement generated 3.20 Tg C in NPP gains, compensating for over 21% of the losses associated with impervious surface expansion, with the Pearl River Delta agglomeration showing the strongest compensation effect (approximately 42%). These findings highlight the potential of urban green infrastructure to mitigate carbon losses and emphasize the importance of embedding multi-scale ecological management into strategies for sustainable urbanization and carbon neutrality.