(366b) Re-Wiring the Domestic Food Trade for Reducing Irrigation Impacts in the United States

The food-energy-water(FEW) nexus concept recognizes the interconnected nature of the three systems and advocates managing them together to avoid sub-optimal outcomes and unintended consequences. The FEW nexus challenges differ for developed agro-economies such as the United States(U.S.) from agriculture centric developing nations. For U.S., irrigation provides a crucial link to study interconnections between food-water-energy resources. Additionally, food trade is an integral part of food system as it bridges the growing disconnect between production and consumption centers. Depending on the origin of food production, trade can potentially reduce environmental impact of food consumption. However, environmental impacts of agriculture are not directly considered when forming trade connections and may carry risk associated with suboptimal FEW system management (e.g. sourcing food from water-depleted regions). Structural optimization of food trade provides an opportunity to assess whether FEW resources can be managed together with respect to land-water-energy constraints. This work focuses on environmental impacts associated with FEW systems, namely- water required for irrigation, greenhouse gas(GHG) emissions from commodity transport, and pumping irrigation water for food production.

We leverage publicly available datasets from agriculture census and transportation surveys to create a network of regional food trade. The study combines data on state-wide food production, irrigation statistics, and life cycle inventory databases to convert food transfer network into embedded irrigation water, and GHG emission flows. Specifically, we focus on rice production and trade in the U.S. Rice is one of the staple food crops and produced entirely through irrigation, making it an ideal crop for the irrigation impact analysis. Majority of rice production is concentrated in four regions-Arkansas Grand Prairie, Mississippi Delta, Gulf Coast, and Sacramento Valley. The model optimizes rice trade network by keeping the demand constant and evaluate a series of constraints pertaining to reduction in energy and water usage to understand whether production can be shifted realistically to optimize all three systems. We limit the potential crop shifting to current production regions to account for soil and climatic parameters conducive to crop production. Additionally, we consider type of rice grown in each region as crop physiological requirements may differ based on the rice type. Preliminary results of the optimization framework reveals a rewired domestic rice trade network with a 27% reduction in virtual water use (15.1 billion m³ reduced to 11.1 billion m³). However, due to shift in production to more water-efficient states, a 10% increase in marginal land use is required. The land results are significant as they demonstrate the resulting sub-optimal solutions when interdependencies are overlooked. On the other hand, minimizing groundwater usage from water stressed areas also reduces overall GHG emissions demonstrating that these are not mutually exclusive goals in many states. The Pareto optimal solutions showing water-GHG emissions tradeoffs resulting from the model will be presented and their implications for regional FEW nexus sustainability will be discussed.