Bayesian networks for environmental flow decision-making and an application in the Yellow River estuary, China

Abstract. We proposed an approach for environmental flow decision-making based on Bayesian networks considering seasonal water use conflicts between agriculture and ecosystems. Three steps were included in the approach: water shortage assessment after environmental flow allocation using a production-loss model considering temporal variations of river flows; trade-off analysis of water use outcomes by Bayesian networks; and environmental flow decision-making based on a risk assessment under different management strategies. An agricultural water shortage model and a production-loss model were integrated after satisfying environmental flows with temporal variability. The case study in the Yellow River estuary indicated that the average difference of acceptable economic loss for winter wheat irrigation stakeholders was 10% between water saving measures and water diversion projects. The combination of water diversion projects and water-saving measures would allow 4.1% more river inflow to be allocated to ecological needs in normal years without further economic losses in agriculture.