CHANGING RESPONSES IN HYDROLOGY - ASSESSING THE UNCERTAINTY IN PHYSICALLY BASED MODEL PREDICTIONS

Due to the large number of model parameters requiring calibration and their inherent uncertainty, the practical application of physically based hydrologic models is not a straightforward task and yet has received inadequate attention in the literature. This work investigates the determination and usefulness of a measure of predictive uncertainty in a particular distributed physically based model, using the methods of Rosenblueth (1975) and Monte Carlo simulation, in an application to an upland catchment in Wales. An examination of the role of predictive uncertainty in assessing the hydrological effect of land use change is also made. The results of the study suggest that, even following parameter constraint through calibration, the predictive uncertainty may be high and can be sensitive to the effects of land use change.