Abstract
"人類文明與經濟的發展，取決於人們利用河川資源與洪害防治的能力。由過去二十年來全球各地一再發生的重大洪水事件，以及所造成土砂沖蝕的增加顯示，在面臨全球氣候變遷情況下，迫切需要水文工程師發展洪水事件與土砂沖蝕之分析方法。然而，適當的水工結構物設計與管理，往往需要詳細的河川流量資料以進行時間序列分析。但是在開發中國家，河川流量資料通常甚少或者完全無紀錄，以致於無法在較易受災的特殊地形與氣候環境中預測水文極端事件。 相對於河川流量資料，降雨資料通常較容易獲得。因此基於地文與雨量資料容易取得之情況下，本計畫主要目的乃在發展可應用於無流量紀錄地區之逕流模式與泥砂運移模式，以模擬不同地文及氣候條件下之非穩態與非線性水文反應系統。於本計畫前期階段，將由印度與台灣雙方學者個別發展分析方法，而後整合雙方的經驗，嘗試發展新的模式，且利用印度及台灣雙方水文資料，以測試所發展模式之可行性。此項印度-台灣合作計畫之主要目標為：a)發展完整的泥砂量預測模式與洪水預報模式，以應用於水文資料缺乏的開發中國家，進行不同重現期水文分析；b)針對無紀錄集水區之水工結構物進行水文與工程經濟合理性評估；c)台印雙方學者相互交流水文紀錄資料與水文模式技術；d)雙方共同進行諮詢、研討會以及訓練課程；(e)發表國內與國際期刊論文。 本計畫各年度主要工作內容如下： 第一年：集水區逕流模式(半分佈式模式) 第一年計畫將蒐集印度及台灣水文紀錄及地文資料，並分析集水區之地文因子如河川坡度，集水面積，河川網路，以及水力幾何參數，以提供台印雙方團隊所發展降雨逕流模式之應用。基於線性系統之假設，可將集水區水文反應假設為地表逕流與地表下逕流之疊加。地表逕流可藉由考慮隨時間變異之單位歷線模式進行計算，應用運動波理論以解析的方式，求解地表逕流區域之漫地流與渠流逕流時間，並利用達西定律計算地表下區域之逕流運行時間(Lee and Chang, 2005)。前述所提及之方式，可利用經驗、物理或概念化方法，配合已廣泛採用之Phi指數法(Gray, 1973)進行計算。由於降雨損失法須依靠真實的降雨逕流及土壤水文資料，因此逕流模擬方式無法應用於無紀錄集水區。然而印方(Kaur et al., 2004)曾提出以USDA之SCS-CN法(SCS, 1972)進行入滲估計，應可克服此模式限制。研究中為正確進行降雨損失推估，將會利用代表性降雨事件檢定曲線值(CN值)，而後進行逕流演算。印方團隊曾發展出利用降雨逕流資料，以計算降雨損失(Kaur and Arekhi, 2008)。因此本計畫將應用由台方所發展之運動波-地貌瞬時單位歷線模式(Lee and Yen, 1997; Yen and Lee, 1997; Lee and Chang, 2005)，並配合印方團隊所發展之降雨損失法，計算超滲降雨(Kaur and Arekhi, 2008)，以模擬集水區逕流。研究中將選定印度及台灣之試驗集水區，以測試本模式之可行性。為避免在不同氣候條件下，以不適當的方式模擬逕流歷程，未來將合併修正之運動波-地貌瞬時單位歷線模式與超滲降雨模式架構，模式參數也將透過詳盡的參數敏感度分析。 第二年：集水區泥砂運移模式(半分佈式模式) 當大量降雨落在坡陡流急之集水區時，常造成嚴重土壤沖蝕情形。由於泥砂之沖蝕及運移過程相當複雜，致使集水區之泥砂濃度並不容易估算。本計畫將改進Yang and Lee (2008)所提出之物理模式方法，以估算暴雨期間之泥砂量。利用集水區土壤與地形特性以及泥砂於不同級序之漫地流區域與渠流運移時間之解析解，以建立地貌瞬時單位泥砂歷線模式(geomorphologic instantaneous unit sedimentgraph, GIUS)，並描述單位瞬時超滲降雨輸入所造成泥砂流出量之時間分佈。地貌瞬時單位泥砂歷線模式可以超滲降雨強度與泥砂運移率之函數表示，因此泥砂流出量即可經由降雨強度及瞬時單位泥砂歷線摺合積分得。研究中將利用印度及台灣試驗集水區資料進行驗證，而泥砂歷線之模式參數敏感度亦可求得。本研究所建立模式可應用於水資源工程設計，以估計泥砂流出量。 第三年：集水區逕流與泥砂運移模式(全分佈式模式) 第三年計畫目的為分析土壤、地形及地表覆蓋空間分佈所造成逕流與泥砂運移特性之影響。因此為達成該目標，將發展以格點為基礎之運動波-地貌瞬時單位歷線，此全分佈式模式將取代先前級序平均之運動波-地貌瞬時單位歷線模式。相同地，台方之泥沙運移模式(Yang and Lee, 2008)與印方之MUSLE模式(Arekhi and Kaur, 2007)將應用於印度與台灣集水區，以產生泥砂歷線。計畫中將以印度及台灣集水區作為模式驗證，由全分佈式模式模擬所得之流量歷線與泥砂歷線(或總輸砂量)將與真實紀錄比對，以測試模式在不同的氣候與地形條件下之可行性。" "Civilization and economic development of society are closely related to the ability to maximize the benefits and minimized the damage caused by the rivers. Frequent occurrence of disastrous floods, during last twenty years, and consequent increased global soil erosions has raised an urgent need to develop new approaches for appropriate designing of hydraulic structures and thereby checking extreme flood and soil loss problems under changing global climatic conditions. Appropriate design and management of hydraulic structures requires detailed time series analysis of stream flow data. However in developing nations, such long term stream flow records are either often too short to predict extreme events or are completely missing for vulnerable geo-climatic setting. In contrast to the stream flow (or discharge) data, the precipitation data is generally more readily available. Thus the present proposal is primarily aimed at developing such user-friendly runoff and sediment transport models, based on readily available geomorphologic and rainfall data, which can be used for simulating impacts of changing geo-climatic conditions on the hydrologic response of any un-gauged/ inadequately gauged non-stationary/ non-linear (watershed) systems. In the initial stage of the project, the appropriate analytical approaches will be developed by both Indian and Taiwanese scientists separately. Thereafter, the experiences from both the sides will be integrated to develop a new approach and the ability of the newly developed models would be tested on the hydrological records from, the mountainous areas of, both India and Taiwan. It is anticipated that the proposed India-Taipei co-operation would go a long way in: a) devising robust approaches for accurate sediment yield estimation and flood forecasting, at different return times, for data sparse developing nations; b) scientific hydro-economic analysis of existing/ proposed conservation structures in any un-gauged or inadequately gauged basins; c) interchange of data and algorithms between Indian and Taiwanese scientists; d) consultations, workshops and probation trainings in both countries; and e) joint domestic and international publications. The proposed work plan of the project is as follows: First year: Watershed runoff transport modeling (semi-distributed model) Hydrological records and geomorphologic data from India and Taiwan will be collected. Watershed geomorphologic factors such as slope, catchment area, channel network and hydraulic geometry parameters will be analyzed and processed through the rainfall-runoff algorithms developed by the Taiwanese and Indian teams. Based on the linear system assumption, the hydrologic response of a watershed would be assumed to be a superposition of the surface and subsurface flows. Surface flows would be accounted by a unit hydrograph based linear module, which explains excess-rainfall (or runoff) variations over time. A kinematic-wave approximation will be used to estimate the mean value of the travel-time probability distributions for runoff in surface-flow regions and channels while Darcy’s law will be adopted to estimate the runoff travel time in subsurface-flow regions (Lee and Chang, 2005). In the above mentioned procedure, abstractions can be accounted through a large number of empirical, physical or conceptual approaches, with the Phi-index method (Gray, 1973) being the most widely applied approach. However, due to the dependency of the aforementioned rainfall abstraction methods on the actual rainfall-runoff and soil hydrologic data, such runoff simulating approaches cannot be applied on un-gauged/ inadequately gauged watersheds. A study by the Indian team (Kaur et al., 2004) illustrated that this limitation can be easily overcome through the USDA’s SCS-CN method (SCS, 1972). However for obtaining accurate rainfall abstractions and hence run-off estimates, careful parameterization of curve number values through calibrations on some representative storms was pre-requisite. Hence, due to the non-availability of event based rainfall-runoff data on the Indian/ Taiwanese watersheds, even this approach was found to be unsuitable for the un-gauged/ in-adequately gauged watersheds. In order to overcome this problem, recently a new algorithm (independent of event based rainfall-runoff data) for computing rainfall abstractions has been developed by the Indian team (Kaur and Arekhi, 2008). Thus, in this project, the Kinematic-Wave-based Geomorphologic Instantaneous Unit Hydrograph model (KW-GIUH) developed by the Taiwan team (Lee and Yen, 1997; Yen and Lee, 1997; Lee and Chang, 2005) will be applied to simulate both surface and subsurface flow processes while the excess rainfall would be accounted by the rainfall abstraction algorithm developed by the Indian team (Kaur and Arekhi, 2008). Sample watersheds from both India and Taiwan will be selected to test the capability of the aforementioned model algorithms. In case of inadequacy of approach to simulate runoff transport processes under different climatic and geomorphologic conditions, further modifications in the KW-GIUH & excess rainfall model structure will be incorporated. A detailed sensitivity analysis of model parameters will also be performed. Second year: Watershed sediment transport modeling (semi-distributed model) Concentrated rainfall usually results in serious soil erosion on steep hill slopes. Since the itinerary of the eroded sediment is complicated, estimating watershed erosion during storms is practically difficult. A preliminary physical model based approach (Yang and Lee, 2008) for sediment yield estimation during storms will be revised in this project. By using soil and watershed geomorphologic information, analytical solutions for sediment travel time in different orders of overland areas and channels will be derived to develop a geomorphologic instantaneous unit sedimentgraph (GIUS) to depict temporal distribution of sediment discharge from an instantaneous excess rainfall input. The resultant GIUS will be a function of the excess rainfall intensity and sediment delivery ratio. The linearity restriction of the unit hydrograph theory can be relaxed. Sediment yields during storm events will be calculated by convoluting rainfall intensities with the proposed GIUS, which will be verified by using field data from sample watersheds in India and Taiwan. Sensitivity of the sedimentgraph to the model parameters will be also investigated. The proposed model may be considered a promising application for sediment yield estimation in the field of water resources design. Third year: Watershed runoff & sediment transport modeling (fully-distributed model) The objective of the third year project is to investigate the effect of the spatial distribution of soil, topography, and land cover on runoff and sediment transport behavior. Thus to achieve this objective, a fully-distributed runoff transport model would be developed by replacing the aforementioned order average KW-GIUH model with a cell/ grid based KW-GIUH model. Simultaneously, sedimentgraphs for the sample Indian/ Taiwanese watersheds would be generated through both the sediment delivery based approach proposed by the Taiwanese team (Yang and Lee, 2008) and the MUSLE model based approach proposed by Indian team (Arekhi and Kaur, 2007). The hydrographs and sedimentgraphs (or total sediment yields) simulated by so developed fully-distributed runoff and sediment transport models would be compared with the observed hydrographs and sedimentgraphs (or total sediment yields) for the sample Indian and Taiwanese watersheds for testing the applicability of the model under different climatic and geomorphologic conditions."
