A novel method for flood forecasting based on the nonlinear computational units cascaded model

作者：Wang Chun Ming; Lin Gwo Fong^*

来源：Paddy and Water Environment, 2015, 13(1): 115-123.

DOI：10.1007/s10333-013-0413-z

• Flood forecasting
• Nonlinear computational units cascaded model
• System memory
• Automated calibration method

To facilitate flood forecasting (FF), a novel method based on the nonlinear computational units cascaded (NCUC) model is proposed in this paper. The proposed method is named the NCUC-FF method. The NCUC-FF method overcomes the shortcomings of two types of methods, respectively, based on the artificial neural networks and the conventional rainfall-runoff models. The NCUC-FF method uses only the rainfall records as input to produce accurate forecasts of runoff, and it has an automated calibration method. These are the two distinctive features of the NCUC-FF method. Furthermore, the NCUC-FF method is flexible to meet the modelers' requirements. Modelers can freely choose the pattern of the NCUC model within the NCUC-FF method. Actual applications of the NCUC-FF method for FF are also presented in this paper. The results show that the performance of the NCUC-FF method is outstanding. It is concluded that the NCUC-FF method is useful for FF.

• 出版日期2015-1