Calibration and evaluation of hydrological models
The calibration of flood forecasting models is important for providing reliable flood forecasts. The thesis will focus on different strategies for calibration and the implications for the robustness of flood simulations.
In this thesis, we will compare models that are calibrated for the entire dataset with models that are calibrated on parts of the dataset, e.g. the highest / lowest water flows and seasons with or without snow.
Questions we want to help answer are, for example: Does a model provide better spring flood forecasts if calibrated only on spring floods?
Is a model that is calibrated on spring floods robust in terms of simulating other types of floods?