Creating skillful and useful water prediction modeling systems remains a challenge. Societal needs require modeling at larger spatial domains (i.e. continental) and unification of freshwater, saltwater, and atmospheric modeling systems. Solving these challenges requires creating new software frameworks, leveraging domain, computer, and data sciences, and using open-source best practices to foster a dynamic community of developers and practitioners. Models operating within these new frameworks must be appropriate with respect to uniqueness of place and physical complexity, ideally utilizing parsimonious formulations that acknowledge uncertainty in processes and parameters. A successful framework promotes model interoperability, allows scientific evaluation of heterogeneous model systems over large areas, and is approachable by domain scientists and engineers, having a low barrier to adding new functionality. This session invites contributions from those making community advances in all aspects of this challenge - freshwater and saltwater domain scientists, computer scientists, software developers, and data scientists.