Modeling data for robust calibration of hydrological model

This dataset contains data used for the calibration of a hydrological model simulating pesticide transfer in a catchment. The simulations are performed using the PESHMELBA model, while accounting for uncertainty in environmental forcing inputs, in particular rainfall time series.. The calibration problem addressed in this dataset considers both model parameters and uncertain forcing inputs. In contrast to classical calibration methods, which rely on a single realization of the forcing conditions, this dataset supports approaches that account for variability in external forcings. The data support the comparison of classical and robust calibration strategies. Robust calibration methods aim to identify parameter values that provide satisfactory model performance across a range of uncertain forcing scenarios, rather than optimizing performance for a single realization. To enable efficient evaluation under uncertainty, stochastic emulation techniques are used to approximate the model response under varying forcing conditions and parameter configurations. This dataset includes: Model input data: parameter sets used for experimental designs of the hydrological model. Simulated outputs: soil moisture profiles generated by the PESHMELBA model under different parameter and forcing configurations. Reference simulations: model outputs used to define calibration objectives and reference scenarios. Calibration and analysis results: outputs related to the evaluation of classical and robust calibration strategies. Methodological results: data used for performance assessment and analysis under uncertainty. Data formats: the dataset is provided in .npy, .RData, and .csv formats. Data reuse: this dataset can be used for studying calibration methods under uncertain forcing conditions, comparing robustness criteria, and analyzing the impact of input uncertainty on hydrological model outputs.