Advances in uncertainty quantification for water resources applications

Engineering and physical sciences often rely on mathematical approximations of reality (aka models) that are parameterized with indirect and noisy observations. Consequently, characterization and quantification of uncertainty is a primary challenge in modern science-based predictions (National Research Council 2012). When applied to water resources, uncertainty quantification (UQ) plays a central role in assessing the reliability and accuracy of hydrological models and making optimal water management decisions based on such models (Tartakovsky 2013). Environmental policy and regulations around the world increasingly require UQ as a basis for the planning of measures able to meet the prescribed objectives and minimize possible hazards for human health and the environment (e.g., EP 2000).