Investigating the Role of Hydrological Model Parameter Uncertainties in Future Streamflow Projections

Abstract

Calibrated hydrological models forced with the climate data from various climate models have been widely employed for future streamflow projection. But a major cause of concern in such an analysis has been the suite of uncertainties inherent in the modeling chain that begins from the climate models and ends with the hydrological models. The uncertainties contributed by the hydrological models have generally been given a lesser focus. In the present research, the contribution of the hydrological model parameter uncertainty has been investigated. The multiobjective evolutionary algorithm (MOEA) is employed for calibrating the hydrological model, the Soil and Water Assessment Tool (SWAT), developed for the Magpie River, located in Northern Ontario. The calibrated model was then forced with the data from an ensemble of six regional climate models for projecting the scenario streamflow and evaluating associated uncertainties. A significant variation in seasonal water availability is projected for the two scenario periods studied. The contribution of the hydrological model parameter uncertainty in the streamflow projection is found to be significant, lying in the range of 16%–83%, depending on the month.