Abstract
Water yield services (WYs) play a crucial role in the hydrological cycle and water resource allocation in terrestrial ecosystems. Therefore, modeling their dynamic variation characteristics and driving mechanisms is of extensive practical significance in guiding ecological management practices in arid and semi-arid regions. Gansu Province is located in the heart of northern China. It is rich in wildlife resources and has numerous ecological reserves, whose ecological transitions profoundly affect the northern region and the ecological security at the national scale. In recent years, Gansu Province has encountered the severe challenge of water resources. It is enduring pollution and a severe imbalance between supply and demand owing to the twofold influence of global warming trends and high-intensity human activities. Based on this, this study quantitatively analyzed the characteristics of the dynamic variation in WYs in Gansu Province using the InVEST model and revealed the key factors driving this dynamic variation. The results show that the WYs in Gansu Province fluctuated between 278.37 and 381.96 mm during 2000-2022, with an average WY of 61.09 mm. The rate of spatial variation in WYs was mainly concentrated between -2 and 5 mm/yr and increased at a rate of 1.41 mm/yr. The spatial heterogeneity of WYs was differentiated significantly by natural and socio-economic influences, with precipitation explaining the highest degree of spatial heterogeneity in WYs (q = 0.49-0.62) and the strongest interaction between precipitation and actual evapotranspiration (q = 0.94). Meanwhile, the interaction between precipitation and land use increased from 0.68 in 2000 to 0.75 in 2022. Moreover, the explanatory power of the interaction between the two showed an increasing trend. In addition, the correlations between each driver and alterations in the WYs showed spatial variations, and the characteristics of each factor differed at different spatial scales. The GDP, proportion of urban construction land, and proportion of arable land area had significant negative spatial effects on WYs. Meanwhile, precipitation had a positive spatial effect on WYs.