2019 Food-Energy-Water Nexus

Integrated Modeling of Food-Energy-Water Systems: Challenges and Opportunities of Quantitative Graphical Networks

Authors

Jeuland, M., Duke University
Borsuk, M. E., Duke University
Olander, L. P., Duke University
Bradbury, K., Duke University
Malof, J. M., Duke University
Food, energy and water systems are under increasing stresses from population growth, development and depletion of natural resources. There is ample modeling capacity for energy-food, water-food and energy-water relationships, but each of these pairwise relationships makes strong assumptions about the others and neglects the role of interacting social and environmental variables. Consequently, interventions in food, energy and water systems have tended to have unpredictable and unintended outcomes. Our work explores the capacity of graphical network approaches to characterize tightly coupled human-natural systems, resolve paradoxes and better anticipate the impacts of policies and investments. We aim to then leverage these approaches into quantitative decision-support tools. We find that evidence synthesis and interpretation can be improved by decomposing high-level associations of interest (e.g., the relationship between crop yield and investments in irrigation) into a graphical structure that accounts for interacting socio-environmental variables. However, the ability to parameterize such detailed models is highly influenced by availability, quality and scale of data. We hypothesize that emerging methods in remote sensing and machine learning can be deployed to address these needs and that the resulting models will improve decision-making in the realm of food, energy and water systems.