Abstract:
Global water models allow us to explore the terrestrial water cycle in an earth-sized digital laboratory to guide science and policy. These models simulate increasingly complex representations of different compartments, such as groundwater, and complex processes, such as seawater intrusion. They are coupled with other models, bringing us closer to a digital twin of Earth and enabling us to carry out complex impact assessments of the far future. At the same time, bias-corrected seasonal forecasts allow us to use these models as a window into the near future. However, global water models remain highly uncertain, and evaluating them remains challenging. This is mainly due to imbalanced data, difficulties in tailoring our models to regionally diverse hydrologic systems, and the need for new model evaluation and calibration methods. Multiple pathways can be taken to reduce these uncertainties by collecting available knowledge in new ways and analyzing this data with new approaches. In this talk, I will outline new strategies for assessing these complex models differently and rigorously and develop more sustainable research software for a more robust science.