Predictions in hydrology rely on the interpretation and synthesis of
data by models. Models integrate several orders of complexity including the geological heterogeneity, the flow and transport processes and the geochemical variability. Models also face the lack of available site data and understanding. Uncertainties stems at every step from the geology and forcing terms characterization to the numerical and mathematical approximations. Uncertainties then propagate to model predictions and to the subsequent risk-assessment studies. Several frameworks have been developed to manage uncertainties like stochastic modeling, multi-modeling and probabilistic risk assessment.

This session is dedicated to recent advances in improving predictions
and understanding their uncertainty. Expected contributions concern non exhaustively the characterization and modeling of uncertainty, the influence of uncertainty on our understanding of hydrologic processes, optimization methods for reducing uncertainty, comparison of different uncertainty modeling methodologies, post-audit of model predictions, integration of additional information, impact of uncertainty on risk assessment and decision making. The objective is to bring together contributions on theoretical aspects, numerical developments and experiences in real field cases.