Analysis of the Water-Food-Energy Nexus and Water Competition Based on a Bayesian Network

Water competition is a key issue in the study of the water-food-energy nexus (WFEN), which can affect water, food, and energy security and can generate notable challenges in water resource management. Since Bayesian network can express parameter uncertainty with a certain probability distribution while reflecting the dependencies of each variable, this study used a Bayesian network to model the WFEN in the Pearl River Region (PRR). The network structure can intuitively represent complex causal relationships, and the form of the probability distribution can effectively reflect the variable uncertainty. The responses of the Bayesian network model under different scenarios were used to analyse the major influencing factors, and water competition relationships in various sectors were explored. The results indicated that water competition between the different sectors was very complex and could dynamically change under the different scenarios. For example, an increase in hydropower and flow to sea could lead to a decrease in irrigation water, but an increase in irrigation water did not necessarily reduce hydropower and flow to sea. Water for hydropower generation and salt tide alleviation were obviously affected by the total offstream water use, but there existed no obvious water competition between these aspects in general. However, when offstream water use remained stable, a competitive relationship was observed between hydropower and flow to sea. Overall, the outcomes of this study could be of great significance to further analyse the WFEN in other regions.