The Uncertainty Analysis Of The Watershed Hydrological Model

Watershed hydrological models illustrate the hydrological process by means of conceptions and abstractions, using mathematical formula,and the equation to describe true hydrological process.when applying the watershed hydrological models for the forecasting and simulating,the hydrological and meteorological data are required to input the models,the oberserved values are required to jude the simulating efficiency, the uncertainties of the watershed hydrological models really exist in the above-metioned process. The influence to the model simulations of the uncertainties from different kinds can be analysed through the uncertainties analysis process,the abundant informatiom to improve the simulating efficiency and the model can also befound .The uncertainties of the watershed hydrological model include the input, the parameters, the model structures and the watershed respones uncertainties. Analysis shows that the model input influences the simulating process and the model parameters selection, bring the uncertainties to the the model simulating process; model parameters have obvious equifinality phenomenon, bring difficulties to choose the appropriate parameter set for the model; The watershed response uncertainties influences the simulating efficiency, and parameters distribution, bring the uncertainties to select the appropriate parameter set.The principle and structure of the GWLF model are introduced ,the input data ,the input parameters and the parameters'scopes are also analyze ed.By analysing the traits of the GWLF model, the soil moisture factor is added to reflect the influence of the moisture condition in order to improve the GWLF model. The principle and the analysis process of the GLUE method is introduced, the the GLUE method is used to analyze the the parameters uncertainties of the GWLF model. Reults shows that the posterior distributions changed a lot compared with the prior distribution, and have obvious higher probabilities interval, the 90% confidence interval of the runoff simulations cover the observed values well.On the basis of the GLUE method, a framework of the integrated uncertainties analysis presented, this framework inlude the model input, the parameters, and watershed response integrated uncertainties. The input uncertainties expressed by means of parameterization, the Gaussian random multiplier is introduced to reflect the input uncertainties. Compared with the GLUE method results, the parameters posterior distributions of the integrated uncertainties changed, the 90% confidence interval of the runoff simulations cover the observed values better.