張斐章臺灣大學：生物環境系統工程學研究所江衍銘Chiang, Yen-MingYen-MingChiang2007-11-272018-06-292007-11-272018-06-292007http://ntur.lib.ntu.edu.tw//handle/246246/56138本文的主要目的為利用雷達及數值天氣預報之訊息整合於類神經網路以有效地建構多階段洪水預測模式。研究藉由類神經網路於降雨-逕流模擬過程之精確性及適用性，針對三項研究主題分別進行探討。首先，應用四組不同數目及變異性之資料分別訓練靜態及動態網路，進而評估其優劣，評比之成果指出，動態網路之輸出誤差普遍來說低於靜態網路，而靜態網路僅能在資料充足且變異性豐富的條件下才可有精確且穩定之預測。 第二研究主題著眼於網路架構之觀點，經分析評估三種不同多階段預測模式之有效性及穩定性，結果顯示運用一序列式傳遞之類神經網路架構可有效提升模式於多階段預測之準確性，此一技巧不僅可提供網路於搜尋過程獲得較佳解之可能性外，更能增加模式預測之可靠性。上述兩項研究成果皆有助於多階段洪水預測模式之建立；為進一步提升模式預測之階段，雨量預報資訊實為不可或缺之訊息，因此，本研究第三項主題為以融合程序有效結合雷達觀測及數值天氣預報模式所繁衍之預報雨量，提高定量降雨預報產品之精確度。 由模式觀點及資料觀點進行評比，可得下述之結論；在定量降雨預報上，本文所提之融合程序驗證其可有效地結合兩組不同之雨量預報資訊並提高未來1至6小時預報雨量之精確度。在多階段洪水預測上，一相當重要之發現為未來1至3小時的水位流量預測，對預報雨量資訊的提供影響有限，洪水預報主要是受前階段之水位流量訊息所影響；反之，模式於4至6小時之預測則高度仰賴雨量預報資訊。總結來說，此研究成果強烈地驗證序列式傳遞架構具有提供準確且穩定之多階段洪水預測能力，而本文所提之融合雨量預報資訊則可進一步提升模式於多階段洪水預測之精確度。The major purpose of this dissertation is to effectively construct artificial neural networks-based multi-step-ahead flood forecasting using radar and numerical weather prediction information. To achieve this goal, three investigations by using neural networks for rainfall estimation and/or rainfall-runoff process simulation have been performed to explore their accuracy and applicability. The first topic investigates the model forecasts through static and dynamic neural networks by using four sets of training data which consist of different sample sizes and contents. Performance of these two types of networks suggest that the dynamic neural network generally could produce better and more stable forecasts than the static neural network, and the static model could produce satisfactory results only when sufficient and adequate training data are provided. The second topic focuses on the evaluation of effectiveness and stability of three neural networks-based multi-step-ahead forecasts in terms of model structures. The results indicate that a neural network with a serial-propagated structure can help in improving the accuracy of forecasts. This concept not only provides a possibility of finding better solution for multi-step-ahead forecasts but enhances the predictive reliability. Results from above two studies are further utilized in the third topic which is to construct a precise and feasible multi-step-ahead flood forecasting. For better multi-step-ahead flood forecasting, there is a necessity to conduct the predicted meteorological information. Therefore, an improved quantitative precipitation forecasting is obtained from a merging procedure that combines radar-derived predictions and precipitation forecasts extracted from a numerical weather prediction model. The comparison of multi-step-ahead flood forecasting derived from the serial- propagated structure and the merged precipitation prediction is made by estimating the timing and the percent error of a predicted peak flow relate to observed peak flow and the corresponding improvement. Based on the comprehensive comparison, the merging procedure successfully demonstrates the capability of efficiently combining the information from both rainfall sources and improves the accuracy of 1-6 h precipitation predictions. For multi-step-ahead flood forecasting, an important finding is the hydrologic responses seem not sensitive to the precipitation predictions in short lead times (in our case 1 to 3 hours) but dominate by previous runoff information, whereas the model forecasts are highly dependent on predicted precipitation information for lead time greater than 3 hours. Overall, the results strongly demonstrate that accurate and stable multi-step-ahead flood forecasting can be obtained from a serial-propagated structure and enhanced by the proposed precipitation predictions.Abstract i 中文摘要 iii Contents v List of Figures vii List of Tables ix Chapter 1 Introduction 1 1.1 Motivation 1 1.2 Structure of the Dissertation 3 Chapter 2 Comparison of Static and Dynamic Neural Networks 5 2.1 Artificial Neural Networks 5 2.2 Static-Feedforward and Dynamic-Feedback Neural Networks 7 2.2.1 Feedforward architecture 7 2.2.2 Feedback architecture 9 2.3 Learning Algorithms 10 2.4 Application 16 2.4.1 Study area 16 2.4.2 Determining an ANN structure 17 2.4.3 Selecting the best static-feedforward neural network structure 20 2.5 Results and Discussion 22 2.6 Summary 27 Chapter 3 Investigation on the Structure for Multi-Step-Ahead Forecasts 29 3.1 Overview of Flood Forecasting 29 3.2 Methodologies for Multi-Step-Ahead Forecasts 31 3.2.1 Importance of multi-step-ahead forecasts 31 3.2.2 The architectures of ANN-based multi-step-ahead forecasts 32 3.3 Application 35 3.3.1 Study area 35 3.3.2 Description of data and ANN model set up 36 3.4 Results and Discussion 40 3.4.1 Experiment I 40 3.4.2 Experiment II 45 3.5 Summary 49 Chapter 4 Multi-Step-Ahead Flood Forecasting 51 4.1 Quantitative Precipitation Forecasting 51 4.1.1 Applications of radar information in QPF 54 4.1.2 Numerical weather prediction model 56 4.2 QPF from Radar and NWP Information 58 4.2.1 Dataset 59 4.2.2 Description of neural network-based QPF models 65 4.3 QPFs for MSA Flood Forecasting 69 4.3.1 QPF by merging radar and NWP precipitation forecasts 69 4.3.2 RNN-based hydrological models 71 4.3.3 Statistics for model assessment 75 4.4 Results of QPF 77 4.5 Results of MSA Flood Forecasting 83 4.6 Summary 99 Chapter 5 Concluding Remarks 101 References 105 Appendix A Curriculum Vitae 113 Appendix B 作者簡歷 1172064020 bytesapplication/pdfen-US類神經網路多階段洪水預測雷達數值天氣預報序列式傳遞架構定量降雨預報融合程序artificial neural networkmulti-step-ahead flood forecastingradarnumerical weather predictionserial-propagated structurequantitative precipitation forecastingmerging procedurethesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/56138/1/ntu-96-D91622006-1.pdf