請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/18881
標題: | 部分現行水利工程實務評估研究 Evaluation of Some Current Hydraulic Engineering Practices |
作者: | Ping-Hung Chen 陳炳宏 |
指導教授: | 蔡丁貴(Ting-Kuei Tsay) |
關鍵字: | 洪水災害,數量排序法,R-largest 方法,全流域不恆定流模式,時域反射器, flood hazard management,basin-wide unsteady flow model,time domain reflectometers ( TDR), |
出版年 : | 2016 |
學位: | 博士 |
摘要: | 本研究基於實務之觀點,對於部分水利工程之分析流程進行完整探討。自洪水災害發生成因分析,再藉由洪水預報模式結果,推求水文、水理等不確定因子影響降雨、逕流主要因素,提出以降低洪水發生機率的概念,以規劃防洪工程。本研究分析降雨資料統計特性及分佈趨勢,提出不同水文頻率分析方法差異比較,另結合採用全流域不恆定流水理分析數值模式演算,估算實際流量,研擬合理堤防出水高度與洪水回歸週期意義,整體可作為技術規範修訂之基礎。此外,因傳統監測技術,受限現地條件,無法適用不同環境,本研究亦提出如何以改良監測技術取得完整水文資料,提高水理演算精確性,研擬適當防洪政策。
本研究架構上從分析以颱風定量降雨量預報模式、降雨逕流水位預報模式與河川洪水位演算模式,探討三個層次模式之不確定性因子影響,進行不同洪水預報模式演算防洪動態數值模擬,藉以預測颱洪時期河川水位,比較各種洪水預報之雨量、河川水位值和實際之雨量、河川水位量測值的誤差,以確認不同洪水預報模式實用性及主要影響洪水災害之不確定因子,探討模式系統誤差及預報過程遭受外界不確定因素影響的誤差原因,以提升洪水預報準確性。其次本研究檢視傳統水文分析年最大值法(annual maximum series)之限制,提出採用數量排序法(quantity ranking series)統計水文資料,結合R-largest方法利用現今雨量分佈趨勢推估未來降雨量。以中央管河川淡水河流域作為研究範圍,比較不恆定流與定量流理水理模式差異,最後再利用時域反射器電磁波監測水位,嘗試解決傳統水位計使用瓶頸,以供模式率定驗證使用。 本研究分析以全流域不恆定流河川洪水演算模式,結果得到洪水預報河川水位及實際河川水位誤差最小,不確定因子降雨量為影響洪水災害主要關鍵因子。水文分析上數量排序法可避免遺漏極端氣候降雨事件,在低重現期時數量排序法比年最大值法保守,利用數量排序法規劃設計可降低低重現期淹水機率。R-largest方法以降雨趨勢分布推估高、低重現期雨量及流量,比過去傳統方法外插歷史資料得到更精準統計分布結果。用R-largest方法推估高重現期流量比傳統方法年最大值低,代表過去高重現期洪水量之推估過於偏大,導致堤防高度設計偏於保守。再以不恆定流水理模式演算,結合設計洪水條件情境模擬結果探討水理特性,透過參數率定及適用性驗證,反應河川真實流況,以檢討堤防合理出水高度及洪水量與回歸週期關係,因地制宜規劃防洪工程。時域反射器( TDR)透過電磁波監測水位,提高監測精度,改善預報時效,解決壓力式水位計無法克服淤積問題及雷達式水位計無法直接在密閉式雨水下水道使用瓶頸,再提供決策單位重新評估水利相關參考手冊是否有改進之處。 本研究彙整現行部分水利工程實務各面向之問題,嘗試建立一改善之分析架構下,用以評估防洪設施保護標準,藉此可以降低不同之不確定性來源,正確瞭解現有架構之問題,以此重新審視防洪工程成本效益,以提出最佳經濟效益之防洪對策,降低洪水災害。 This study is based on practical perspectives to examine different issues in the analysis of part of hydraulic engineering. The author investigates topics from the causes of flood hazard, and the flood forecast result, to determine the main factor of uncertainty in precipitation-runoff process. By this way, this study purposes a concept to reduce the probability of flood occurrences. The author analyzes the participation statistical characteristics and distribution trend to provide the modification of technical manual. Using unsteady hydraulic model to route and calculate real flowrate, and determine the reasonable water stage with freeboard. This study also examines the improvement of current measurement equipment. By this way, we have to increase the accuracy of flood forecast and improve the calculation precision of hydraulic simulation technique and determine the appropriate strategy. The framework this study investigated is form the typhoon-precipitation quantity forecast model, precipitation-runoff model and flood routing model to explore the different uncertain factor. Then this study performs the flood forecast model to do the dynamic river stage numerical simulation to predict the river stage during flood. Therefore, from the difference between real measure water stage and simulated value, the practicability of model and other uncertain factors can be determined. It was used to estimate the model error in different stage, and understand to external uncertainty factors in modelling. By this way, the accuracy of prediction could be level up. Next It is suggested to using quantity ranking method to do the frequency analysis rather than limited annual maximum series. R-largest method uses current participation distribution to determine the future rainfall condition. The region of this study was based on the Tamsui River, governed by the central government. The hydrological data were re-analyzed and used for unsteady hydraulic simulation. This study also suggested to use the electromagnetic waves of time domain reflectometers (TDR) to monitor water levels to enhance the accuracy of monitoring data. Comparing to the limitation of traditional method, it can be more useful for model calibration and validation. This study found the basin-wide unsteady flow model can simulate every point in a basin so the result shows less difference with the observation data. The precipitation is recognized as the main factor of flood. The quantity ranking method we named is the best method which can avoid climate change. With the low return period condition, the quantity ranking method is more conservative than annual maximum method. Using R-largest to estimate the high return period, the flowrate will be larger than traditional method. Combining with unsteady hydraulic routing model in simulating the design storm events, and calibrating the model and discuss the feasibility, the real condition of river can be reflected. Using unsteady hydraulic model to route and calculate real flowrate, and determine the reasonable water stage with freeboard. The electromagnetic waves of time domain reflectometers (TDR) can be used to monitor water levels to enhance the accuracy of monitoring data, improve forecast efficiency, and correct errors. It also can solve the critical deposition problem for traditional technique. The results in the study provide some information to decision-making units for re-assessment to see whether there should be an improvement of engineering manual and technologies on water resources. Thus study summaries the different challenges in part of practical hydraulic engineering. The author tries to apply this improved framework to re-examine the flood protection standard and reduce the uncertainty in design and operation. By understanding the issues of current framework, it can review the benefit and cost of flood prevention measures, purpose the best strategy to reduce the threat of flood. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/18881 |
DOI: | 10.6342/NTU201603604 |
全文授權: | 未授權 |
顯示於系所單位: | 土木工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-105-1.pdf 目前未授權公開取用 | 14.83 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。