MODELS SUPPORTING GROUNDWATER AVAILABILITY DECISIONS: APPROACHES AND METHODS FROM THE USGS MISSISSIPPI ALLUVIAL PLAIN PROJECT (Invited Presentation)

Modern water availability questions require high-level and stakeholder-accepted science support, including numerical models. Moreover, as stakeholders have grown accustomed to near instant access to information such as real-time weather forecasts, methods for construction and refinements of water-resource support systems have not kept pace. Local stakeholders initiated the US Geological Survey's Mississippi Alluvial Plain (MAP) project in 2016, and identified the need to formulate a new approach for applying models to decision support. In the MAP, a variety of physically based and machine learning models are used, in a number of configurations, depending on the societal question being asked. The overarching goal is to provide responsive, consistent, and insightful high-quality science even as the decision-making concerns and associated forecasts change. Challenges to this vision include well-recognized problems of system scale and shortcomings in our ability to characterize an unseen subsurface, but also include decision making issues of timely simulation of moving forecast targets, ensuring scientific reproducibility, and developing a common language across a large, multidisciplinary team of scientists and stakeholders. Upon completion, the MAP project will have developed a range of capabilities (e.g., incorporation of airborne geophysical data, automated inset model construction, socioeconomic linkages to groundwater models) and interdisciplinary workflows. Given the widespread occurrence of similar water availability issues, the MAP work also includes efforts to enhance transferability beyond the MAP area, such as documentation of tools, scripts, and online code repositories.