Interactive loops between anthropogenic urban-rural construction and natural systems in China: quantifying green development pathways and policy effectiveness

Introduction Accelerating urbanization and rural construction (URC) in China profoundly alter natural systems (land cover, ecosystems, climate regulation). The existing green development index (GDI) systems remain inadequate in capturing anthropogenic impacts on natural ecosystems and lack an integrated framework that combines remote sensing and statistical data. Using these combined parameters, this study examines bidirectional interactions between anthropogenic construction activities and natural processes. Methods We calibrate the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS)-Entropy Weight calculation to measure the value of 286 cities or higher than the prefecture level from 2010 to 2023. The GD level of URC in China in 2023 is further by the clustering method. Finally, the deep Q-network (DQN) model is developed to simulate the policy responses under two scenarios of nature-society interactions. Results Results show that the standardized GDI of URC increases from 0.404 in 2010 to 0.553 in 2023. Four clusters are identified, including Advantaged (29 cities), Second best (68 cities), Average (120 cities), and Weak (69 cities). DQN simulations indicate that increasing renewable energy by 20% improves GDI more than a 5% reduction in emissions. The renewable energy pathway is preferred in over 94% of episodes across sensitivity tests. Discussion The study recommends scaling renewable energy integration in Average and Weak clusters, while Advantaged cities play roles in carbon-neutral practices.