Multiscale analysis framework for the Iowa Water-Energy-Food nexus
This research explored Iowa’s Water-Energy-Food (WEF) resource system by modeling with Markov random field and analyzing with network analysis. With recognizing the gap between nexus modeling and communication in WEF resource management, the purpose of this research was to close the gap by proposing a framework of modeling and analyzing heterogeneous data in the WEF nexus discipline. The proposed framework aimed to discover interlinkage and characterize structural patterns between domains of Iowa’s WEF resource system. The biophysical and economic data were collected from multiple sources and processed with a standardized data aggregation pipeline. The first objective of this research was focused on the modeling method. The goal for this part of the research is to determine the technique to model Iowa’s WEF resource system. The appropriate technique helps to identify the interlinkages between components with minimal subjectivity and closes the gap of communication via intuitive visualizations. We considered the model by the data availability and the capabilities of modeling and direct visualization at different scale. As a result, we proposed the method of coupling the data aggregation pipeline with the probabilistic graphical model that uses the same scheme to model and visualize a large-scale system at different scales. The second objective of this research was focused on the characterization of Iowa’s WEF resource system. The goal for this part of the research is to identify the interlinkages and structural patterns of the system across multiple spatiotemporal scales. The multiscale analysis grasped the characteristics of system at finer levels and connected the understanding of the behaviors of the overall system. Betweenness centrality, associativity coefficient, and degree distribution were applied to investigate the characteristics of the models across different scales. As a result, we identify Iowa’s WEF system is a free-scale, disassortative network that well-connected components would less likely connect to each other and more likely connect to the less-connected components. The analysis also suggested that the hydrologic responses was crucial in Iowa’s WEF resource system.