Supporting Reasoning about Future Water Scenarios through the Sustainable Water through Integrated Modeling (SWIM) Framework

SWIM is a human-technology framework that supports the exploration of future water trajectories given plausible scenarios of climate, population, agriculture and technology. It is designed to enable stakeholders and decision-makers to utilize rigorous scientific models that they otherwise might have difficulty understanding. SWIM has an overarching research objective of integrating human processes and technical functionality to enable the co-generation of narratives scientific stories interpreted from the output data generated by any given selection of inputs. The narratives are partly generated by the system using semantic techniques and partly generated by the human using the system as they reason with the data and scientific models. SWIM has three key technical components: 1) Semantic Web techniques for describing data and models using knowledge graphs, recommending model outcomes (AI), capturing the provenance of information, and generating technical narratives and visualizations; 2) a state-of-the-art backend that enables rapid incorporation of scientific models and exchange of information between models, and 3) a human-centered Web-based interface (SWIM 2.0) designed to make data and scientific models transparent to non-experts. SWIM supports two key human processes: 1) human cognition and reasoning about data into the future; and 2) human communication about water resources. SWIM has been under development for five years and has undergone testing of both the usefulness of the approach and the usability of the interface, with positive results. This presentation will describe SWIM, its technical components, human processes that leverage the technology, and testing approach.