How can multiscenario flow paths of water supply services be simulated? A supply-flow-demand model of ecosystem services across a typical basin in China

Ecosystems provide many benefits to humans, and among them, water supply is crucial for human survival and development. This research focused on the Yangtze River Basin as the research area, quantitatively evaluated the temporal-spatial dynamic changes in the supply and demand of water supply services and determined the spatial relationship between the supply and demand regions of water supply services. We constructed the supply-flow-demand model of water supply service to quantify its flow. In our research, the Bayesian model was used to establish a multiscenario model of the water supply service flow path to simulate it and clarify its spatial flow path, flow direction and flow magnitude from the supply region to the demand region and determine its changing characteristics and driving factors in the basin. The results show that (1) In 2010, 2015 and 2020, the amount of water supply services showed a decreasing trend and was approximately 133.57 × 1012 m3, 129.97 × 1012 m3 and 120.82 × 1012 m3, respectively. (2) From 2010 to 2020, the trend of the cumulative flow of water supply service flow decreased each year and was 59.814 × 1012 m3, 56.930 × 1012 m3, 56.325 × 1012 m3 respectively. (3) Under the multiscenario simulation, the flow path of the water supply service was generally the same. The proportion of the water supply region was the highest under the green environmental protection scenario, at 73.8 %, and the proportion of the water demand region was the highest under the economic development and social progress scenario, at 27.3 %. (4) The provinces and municipalities in the basin were divided into three types of regions according to the matching relationship between supply and demand: catchment region, flow pass-through region and outflow region. The number of outflow regions was lowest, accounting for 23.53 %% of the regions, while the number of flow pass-through regions was the highest, accounting for 52.94 %.