王如意2006-07-252018-06-292006-07-252018-06-292001http://ntur.lib.ntu.edu.tw//handle/246246/10687本研究之目的為建立一種機制，期使碎形之理 論可實際應用於集水區之逕流推估上。一般碎形河 川網路生成之方法乃利用碎形生成單元(fractal generator)經遞迴疊代(recursive replacement)之方法 繁衍合成，本研究並進一步引用寬度函數(width function)詮釋集水區之河川網路結構，進而利用歸 納之方法推導出適合各種碎形生成單元類型之寬 度函數通式以繁衍合成碎形河川網路，並代入寬度 函數為主軸之地貌型瞬時單位歷線(width-function based geomorphologic instantaneous unit hydrograph, WF-GIUH)模式以進行集水區之降雨－逕流推估。 研究中並以河川網路信息熵(informational entropy) 與累積寬度函數(cumulative width function, CWF)作 為選擇合適碎形生成單元之判別標準。本研究選用 臺灣北部淡水河流域以及中部大甲溪流域之德基 水庫上游集水區數場颱洪事件作為模式檢定與驗 證以評估模式之適用性。由研究結果顯示，本研究 已成功地將碎形理論實際應用於集水區之逕流推 估應用上，且該機制可進一步加以研析以期能適用 於中上游集水區之颱洪事件逕流模擬，俾供臺灣集 水區防洪規劃及水土保持研析之參考應用。An approach by applying the fractal concept to estimate hydrologic response is proposed in this study by matching suitable self-similar networks (SSNs) to a specific watershed and modeling the runoff with a width-function based geomorphologic instantaneous unit hydrograph (WF-GIUH). In order to work out the identification between a specific basin and SSNs that are generated by generators, a generalized width function is derived. Subsequently, cumulative width functions ( CWF) on the basis of the derived function, as well as the informational entropies are used as criteria to decide the best pattern of the generator for the specific watershed. The WF-GIUH model is then applied to calculate the runoff of the project watershed as an outcome of the estimation. To assess the adaptability of the estimation model, several watersheds of northern and middle Taiwan are selected as study areas. The analytical results of the outflow estimation indicate that the fractal algorithm can be implemented successfully for the calculation of hydrologic responses.application/pdf145959 bytesapplication/pdfzh-TW國立臺灣大學生物環境系統工程學系暨研究所碎形生成單元寬度函數信息熵Fractal generatorWidth functionInformational entropyreporthttp://ntur.lib.ntu.edu.tw/bitstream/246246/10687/1/892211E002143.pdf