Analyzing the driving mechanisms of grain virtual water flow based on the case of China's main grains

To define the complicated driving mechanisms of virtual water flow (VWF) in China's main grains production, this study firstly uses the entropy method and a K-means clustering approach to classify the driving types of VWF based on the data from 31 provinces in the mainland of China, and then applies the PLS structural equation model to analyze the driving mechanisms of VWF in four representative provinces or municipalities(Zhejiang, Heilongjiang, Xinjiang and Shanghai)selected by the classified driving types. The results show that the driving types of VWF can be divided into four types, namely, ecology-economy, ecology-environment, ecology-water use, and comprehensive type. In the four typical regions at provincial level, the social factors directly affect the VWF, and reversely it has a direct impact on the nature, environment and economy of the above four regions. In particular, the water use has a direct impact on the VWF in Heilongjiang attached to the ecologic-environmental type. The findings of this paper manifest that the social indicators rather than the indicators in water use play an important role in the VWF, that is, our findings further testify the status quo of China's VWF: the regions with rich water resources import the virtual water while the regions with poor water resources export the virtual water. Finally, for promoting the rational flow of virtual water we propose some corresponding adjustment countermeasures and suggestions that provide a theoretical basis for the rational allocation and management of water resources.