THE MODEL IS NOT THE DESTINATION: A 10 POINT CHECKLIST FOR GETTING MODELS OFF THE SHELF AND INTO PRACTICE WITH PARTICIPATORY PROCESSES AND ARTIFICIAL INTELLIGENCE (Invited Presentation)

Scientific models provide a means to understand and interact with complex systems, such as groundwater. While technological tools and techniques for groundwater modelling have advanced steadily over several decades, the use and adoption of modeled information into regular management policy and practice has stuttered. Despite the promise of decision support and integrated groundwater models, advanced information systems that are regularly used, and useful, to people usually remain out of reach. If science alone isn’t enough, then what factors and approaches can accelerate adoption of knowledge?

Looking at barriers to adoption and uptake of modeled information, there are two areas of emergent research that provide pathways to rapid improvements: Artificial Intelligence (AI) and Participatory Modeling (PM). These two fields augment science models in areas that are otherwise ignored or treated as ancillary by most modellers. AI provides gains by assisting with approaches to balance tradeoffs and suggest optimal, or ‘satisficing’, solutions across multiple attributes of concern. While AI can support humans reason about complex problems, yet logic and reason cannot change the mind or position of a stubborn stakeholder. For that reason, scientists need to learn PM skills. PM provides approaches to co-design with communities and decision makers. Connecting credible science models with engaged people leads uptake and implementation of real solutions. This presentation draws from numerous integrated groundwater case studies where scientific models formed the core of a decision support process and leveraged both AI and PM approaches. A set of key actions that a modeler can take to improve outcomes and avoid critical barriers is shared. Computer-based models provide evidence-based insight about aquifer system response and behavior. In general, the solutions lie in the synthesis across science, people, and practice. AI and PM provide concrete actions and tools that scientists can employ to expand the usefulness and influence of models. Models offer the promise of science-informed decision making, but building a credible model is not the destination. In an expanded context, AI and PM models allow society to embark on a journey towards improved choices in actual practice.