Leveraging AI, Large Language Models, and Co-Authorship Network Visualization to Globally Understand the Water-Energy-Food-Health Nexus

Jordan is one of the most water-scarce regions in the world, facing climate change impacts on water, energy, and food—the core components of the WEF Nexus. Health, as an additional dimension of the nexus, is being investigated through the NIH-funded Global Center on Climate Change, Water, Energy, Food, and Health Systems (GC3WEFH). A key component of the Center is its Data Hub, which focuses on providing analytical access to datasets that reflect the WEFH nexus components and assembling an open-source software ecosystem to support integrative research while adhering to FAIR (Findable, Accessible, Interoperable, Reusable) principles.

This presentation demonstrates how large language models (LLMs) are transforming our ability to explore the complex interdependencies within the WEFH Nexus. By extracting insights from interdisciplinary sources—such as scientific articles, policy documents, and environmental and health datasets—LLMs provide powerful tools for integrated data analysis and decision-making across these critical domains. Built on the SuAVE (Survey Analysis via Visual Exploration, suave.sdsc.edu) platform, the Data Hub catalog enables intuitive browsing, querying, and faceted searches of Jordan-specific datasets. To enhance accessibility, LLM-based applications are integrated into the hub, allowing natural language queries to generate tables, maps, and visualizations, revealing interrelationships among nexus indicators such as the effects of climate change on water quality and health outcomes. Additional tools evaluate the AI-readiness of datasets and implement strategies to improve their usability for machine learning applications. These innovations enable deeper insights into the WEFH Nexus, supporting simulations of system sustainability and assessing the health impacts of water, food, and energy-focused strategies in environmentally stressed regions.

To further understand the global research landscape of the nexus, we constructed and analyzed a global co-authorship network of research articles referencing all four nexus components in their titles or abstracts. Using OpenAlex, an open-access bibliographic database, and the network analysis extension of the SuAVE platform, we visualized and examined the evolution of research collaborations, emerging topics, and knowledge gaps. Our analysis revealed that over 60% of WEFH-related publications have been produced in the last four years, reflecting a rapidly expanding but still fragmented field. The co-authorship network exhibits higher clustering and fragmentation compared to more established research areas, such as the Water-Energy-Food Nexus, which is characteristic of emerging disciplines. Key topics identified within the WEFH Nexus emphasize sanitation, water quality, and water treatment (water); wellness, safety, and public health systems (health); crop yields, food security, and nutrition (food); and renewable energy and emissions reduction (energy).

While the United States leads global contributions, accounting for nearly 30% of publications in the field, significant opportunities remain to foster stronger global collaborations and reduce fragmentation in the network. The GC3WEFH is leading this effort through a multi-institutional, international collaboration focused on modeling the climate impacts on vulnerable communities in water-scarce areas of Jordan.

This work is supported by the US National Institutes of Health, Fogarty International Center, under award # 1P20TW012709-01.