主要农产品虚拟水贸易网络稳定性仿真模拟及优化调控研究

Virtual water trade can effectively alleviate the pressure on water resources and insure the food security, but there are limitations, namely regional applicability. Because there is the mismatch between the virtual water trade patterns of the major agricultural products in China, that lead to some risks in the virtual water trade of the main agricultural products. In this study, the stability of the virtual water trade network is discriminated, by constructing the virtual water trade flows relationship matrix of the Chinese interregional major agricultural products from 2001 to 2015, introducing the concept of ecological network and small world network to predict the presence of potential risk of virtual water trade network, measuring the virtual water trade network survivability. Using factor analysis, regression analysis and quantitative research methods such as structural equation model, we try to identify the key factors affecting the stability of the network virtual water trade. On this basis, the effects and the mechanism of action of various factors are revealed by the integrated use of visual-PLS model and the BP-DEMATEL model. We try to establish the system dynamics flow diagram, simulate and predict the evolution trend of the network stability of virtual water trade. The adjustment programs of the network stability of virtual water trade should be evaluated and selected by the projection pursuit evaluation model. Finally, we try to put forward specific countermeasures and suggestions to adjust the virtual water trade mode and path of China's major agricultural products, in order to provide scientific basis for the rational allocation of water resources in China.

虚拟水贸易可以有效缓解水资源压力并保障粮食安全，但却存在区域适用性问题。由于我国主要农产品虚拟水贸易格局与水资源分布存在错位现象，导致目前我国主要农产品虚拟水贸易存在一定风险。本研究通过构建2001-2015年中国区际间主要农产品虚拟水贸易流量关系矩阵，引入生态网络及小世界网络等相关概念，预测虚拟水贸易网络存在的潜在风险，测度虚拟水贸易网络的抗毁性，判别虚拟水贸易网络的稳定性。运用因子分析、回归分析和结构方程模型等定量研究方法，识别影响虚拟水贸易网络稳定性的关键因素。在此基础上综合运用visual-PLS模型和BP-DEMATEL等模型揭示各因素的影响方式及作用机理。建立系统动力学流图，仿真预测虚拟水贸易网络稳定性的演变趋势。运用投影寻踪评价模型，对虚拟水贸易网络稳定性调整方案进行评价和优选。最后提出调整我国主要农产品虚拟水贸易方式和路径的具体对策和建议。

我国农业生产受到耕地和水资源的严重限制，水资源短缺和粮食危机已成为制约我国经济发展的主要因素。虚拟水贸易研究立足于缓解水资源压力并保障粮食安全。本项目以中国31个省、市、自治区为研究区域，选取水稻、玉米和小麦为研究对象，系统研究了我国虚拟水贸易网络的稳定性和可持续性、虚拟水贸易网络的演变趋势和网络稳定性演化趋势，以及虚拟水贸易网络风险预测。主要研究内容和结构如下：（1）清晰刻画了虚拟水贸易网络节点（即31省市自治区）和网络整体特征，中国主要农产品虚拟水贸易网络中节点入度较大的地区集中分布在经济较为发达地区，出度较大的集中在主要粮食供应地区；地区间节点强度差异性大，节点度和强度显示高可变性，表明网络结构极为脆弱；节点度相关性具有异配特征，指示虚拟水流动使各地联系愈加密切的趋势；但网络整体稳定性下降，抗干扰和缓冲能力增强，并寻找到随机攻击和蓄意攻击下的脆弱节点（网络中占主导地位的地区）。以上结果揭示了中国存在潜在的虚拟水供需矛盾。（2）运用相关分析、熵值法、系统聚类及PLS结构方程等方法对中国主要粮食虚拟水贸易驱动机制进行定性与定量分析，中国省际间虚拟水流动驱动因子因地而异，主要有生态经济型、生态环境型、生态用水型和综合型4类种驱动类型；其中，社会因素主导中国主要粮食虚拟水流动，而水的直接驱动影响关系只在个别省份得以体现，且对用水的间接负向影响已十分凸显；虚拟耕地、虚拟农药／化肥及气象要素也对虚拟水贸易效应有着制约作用。（3）根据生态网络、计算机复杂网络方法及网络驱动机制研究，初步阐述了我国虚拟水贸易网络趋向稳定的演化路径，针对地区而言，上海是相对最为脆弱且容易出现水危机的城市，浙江、安徽等12城市也应给予高度重视。从影响因素角度分析，水资源要素并非首要因素，社会因素最为重要。综上，提出引导和促进我国主要农产品虚拟水贸易网络稳定的对策及建议，为相关部门的政策制定和决策提供参考。