In Environmental management
Urban growth has caused environmental problems around the world and profoundly altered the terrestrial carbon cycle, especially net primary productivity (NPP). Sustainable urban development requires a better understanding of the impacts of urban growth on ecosystems. We selected Guangzhou City to analyze the impacts of urban development processes and urban geographic changes on NPP, as well as the correlation between urbanization intensity and NPP, using a deep-learning urbanization characteristic index (UCI). The results showed that the NPP in the study area had clear spatial heterogeneity and declined overall from 2001 to 2013. Guangzhou's urbanization became more and more intense, the mean UCI increased significantly from 0.1293 in 2001 to 0.2879 in 2013, and urban geographic type was dominated by urban exurbs in 2001 and 2013 while urban fringe areas increased most significantly and about 2,320.24 km2 of urban exurbs were converted to urban fringes. There was a significant negative correlation between UCI and NPP in 2001 and 2013, implying that NPP had been negatively influenced by the increasing urban development intensity. The transition of urban exurbs to urban fringes was associated with the highest NPP losses, which was caused by cropland loss and built-up land expansion.
Wu Yanyan, Wu Zhifeng, Liu Xingnan
Deep learning, Guangzhou, Net primary productivity, Urban growth, Urbanization geographic types