Study On Parameter Calibration And Uncertainty Assessment Of Hydrologic Model

During the past 60 years, hydrologic models have been fully studied and widely used in flood forecasting, planning, water resources management, etc, providing the fundamental basis for important decision problems such as flood control, drought resisting, and water resources utilization. One challenge that faces hydrologists in water resources planning is to predict the catchment's response to a given rainfall. However, there are great uncertainty in parameter calibration and flood forecasting because of the complexity of hydrologic processes and the existence of errors of historal data and model structure. In order to solve such problems effectively, and consequently improve the precision of flood forecating and reduce loss of floods or droughts, the paper does in-depth studies on optimization and uncertainty assessment of hydrologic model parameters. The major contents and research progress are as follows:(1) Practical experience with the calibration of hydrologic models suggests that any single-objective function, no matter how carefully chosen, is often inadequate to properly measure all characteristics of the observed data deemed to be important. In this paper, a fuzzy multi-objective SCE-UA (FMOSCE-UA) optimization method for rainfall-runoff models which combines advantages of Pareto ranking and fuzzy multi-objective optimization is developed to solve the multi-objective optimization problem for hydrologic models. The FMOSCE-UA considers different aspects of the hydrograph, such as water balance, overall shape of the hydrograph, peak flow and its corresponding time, which makes the model's behavior match the observed hydrograph more closely. The features and capabilities of FMOSCE-UA are illustrated by means of a simple calibration study for Shuangpai Reservior with Xinanjiang model.(2) A new Markov Chain Monte Carlo (MCMC) method entitled the Parallel Adaptive Metropolis (PAM) algorithm is developed to assess the parameter uncertainty of Xinanjiang model using hydrologic data with 3h temporal interval from Shuangpai Reservoir. The PAM algorithm provides an adaptive MCMC sampler to estimate the posterior probability distribution of parameters under Bayesian framework. The performance of the PAM algorithm is greatly improved in manner of parallel computing. A case study demonstrates that there is considerable uncertainty about the Xinanjiang model parameters. The 90% hydrograph predictionuncertainty ranges associated with the posterior distribution of the parameters estimates can bracket the observed flows well, but not large, indicating that the method is feasible.(3) Based on the Bayesian forecasting system (BFS) framework, a new prior density and likelihood function model is developed with BP artificial neural network (ANN) to study the hydrologic uncertainty of Shuangpai Reservior. Markov chain Monte Carlo (MCMC) method is employed to solve the posterior distribution and statistics of reservoir stage. The model is studied with historic floods of Shuangpai Reservior. The results show that Bayesian probabilistic forecasting model based on BP ANN not only increases forecasting precision greatly but also offers more information for flood control, which makes it possible for decision makers consider the uncertainty of hydrologic forecasting during decision-making and estimate risks of different decisions quantitatively.(4) Traditional flood forecasting and operation of reservoirs in China are based on manual calculations by hydrologists or through standalone computer programs. The main drawbacks of these methods are shortage of prediction time due to time-consuming nature, individual knowledge, lack of communication, absence of experts, etc. A Web-based flood forecasting system (WFFS) is developed on J2EE platform using JBuilder Integrated Development Environment and ORACLE Database. The paper focuses on the key techniques such as design of architecture, abstracting of hydrologic models and Entity-Relation model of database on the conditions of multi-reservoir, multi-model and multi-user. The WFFS brings significant convenience to personnel engaged in flood forecasting and control and allows real-time contribution of a wide range of experts at other spatial locations in times of emergency. A case study demonstrates that multi-tiered architecture offers great flexibility, portability, reusability and reliability. The WFFS is applicable to conditions of multi-reservoir, multi-model and multiuser.Finally, a summary is given and some problems to be further studied are discussed.