Deterministic modeling and uncertainty quantification of wind waves in Ilha Solteira Reservoir, Brazil

Wind waves in reservoirs represent a key hydrodynamic process influencing shoreline stability, navigation safety, and the design of hydraulic infrastructure. Despite their practical relevance, wave prediction in inland waters remains subject to significant uncertainties, particularly related to wind forcing and empirical model parameters. This study integrated deterministic and probabilistic approaches for predicting wind waves in reservoirs. Using a deterministic approach, the Simulating Waves Nearshore (SWAN) model was applied to estimate wave height and period. Key variables analyzed included wind velocity, wind direction, the Joint North Sea Wave Project (JONSWAP) bottom friction coefficient, the whitecapping coefficient, and the depth-induced breaking index. Through a probabilistic approach, uncertainties were quantified using polynomial chaos expansion (PCE), and sensitivity analysis was performed via Sobol indices. This framework was applied to a case study of the Tietê—Paraná Waterway in the Ilha Solteira Reservoir, São Paulo, Brazil. Simulations using the Janssen formulation yielded the most accurate wave height estimates. Sensitivity analysis based on Sobol indices identified wind velocity and the whitecapping coefficient as the most influential factors governing wave behavior. This integrated approach enables the generation of contour maps for wave height and period, offering valuable insights for project planning. Thus, the combination of deterministic and probabilistic analyses enhances the understanding of wind wave dynamics in inland waters.