沖積與軟岩混合河川之動態穩定河槽研究

Study on Dynamic Stable Channel in a Mixed Alluvial and Soft-bedrock River

吳國維(Kuo-Wei Wu)

指導教授 ： 葉克家

國立交通大學/工學院/土木工程系所/博士(2019年)

摘要

臺灣在水資源有限情況下，常需在河道建造攔河堰取水，而橫向建造物之興建往往改變下游水砂條件。加上1999年9月21日由車籠埔斷層及大茅埔-雙冬斷層活動所引起之集集地震，對臺灣中部河川造成大規模河床差異隆起後，加速了河槽幾何型態及縱坡降調整。在自然環境及人為營力交互作用下，泥砂供給不平衡及河道沖刷遷急點移動主導了河道劇烈沖刷現象，既有河床沖積材料大量流失造成河床岩盤裸露，演變成複雜之沖積與軟岩混合之河槽型態，而裸露岩床則因年代較為年輕且易受風化影響，依材料強度特性被歸類為軟弱岩盤。軟岩河槽在歷經數年颱洪事件後最大沖蝕深度可達十公尺以上，已危及鄰近跨河及防洪建造物安全，對於沖積與軟岩混合河川而言，劇烈之軟岩沖蝕常常伴隨著下游沖積河道槽擺盪幅度加大，複雜的河相演變特性常常造成河川管理及治理措施的風險。
為能掌握軟岩及沖積混合河川之河性動態演變特性，避免工程管理潛在問題，本研究首先在二維數值模式(SRH-2D)既有功能下，進行包括沖積河槽岸壁退縮模組、軟岩垂向沖蝕模組及軟岩河槽河岸側向沖蝕等模組擴充，以同時具備計算軟岩河槽及沖積河槽混合河川空間及時間上之水理及沖淤變化之能力。然後，配合歷年觀測資料，以新開發之數值模式進行不同空間尺度(河段、區段及斷面)下之混合河川歷年河槽型態及水理特性分析，並利用臺灣河槽水力幾何特性及河槽穩定坡降經驗公式評估比較，研擬混合河川動態穩定河槽縱坡降。接著，參考國內常用計畫河寬及濁水溪穩定河寬經驗公式外，並以數值方法進行二維動床數值模式模擬、河槽寬度情境分析及臺灣河川特性河槽幾何型態分析等方法，分析研究河段在水、砂及人為營力不變下之最適合動態穩定河寬。最後，結合研擬之動態穩定河槽縱坡降及最適合動態穩定河寬，以新開發之二維動床數值模式進行情境模擬分析，檢核動態穩定河槽效果，供河道穩定工程方案規劃及設計參考。
依據上述分析方法，擬定濁水溪研究河段動態穩定河槽如下：(一)動態穩定縱坡降：依河床材料組成特性及河槽沖蝕行為將研究河段分為四個中尺度區段，分別為向源侵蝕、軟岩沖蝕、混合過渡及沖積區段，各區段依據其獨有之河床材料特性呈現不同之河槽縱坡降變化規則。向源侵蝕區段坡度因應集集攔河堰構造物之保全，縱坡降有持續增加趨勢；軟岩沖蝕及沖積河槽區段分別大致維持0.0058及0.0068動態穩定坡降；而連結軟岩沖蝕及沖積河槽區段之混合過渡區段則有逐漸變緩之趨勢。(二)動態穩定河寬：集集攔河堰至斷面115向源侵蝕區段為350公尺，在軟岩沖蝕區段(斷面108～115)為420公尺，而下游沖積區段為550公尺，至於軟岩與沖積河槽間之混合過渡區段(斷面105～108)間則為450公尺。
本研究主要成果為以自行開發之數值模式，評估混合河川動態穩定河槽變化，創新之處包括(一)開發新的沖積與軟岩混合河川數值模式，使其具備同時模擬軟岩與沖積河槽空間及時間變化之能力，模式功能驗證結果顯示：新開發之數值模式除可用於混合河川水理特性分析外，並可應用於複雜之河川河槽演變情境分析。(二)以不同空間尺度評估混合河川歷年河槽縱坡降變遷化及未來演變趨勢，發現中空間尺度(區段)，可掌握混合河川不同河床特性區段之河槽縱坡降演變特性，為最適合應用於沖積與軟岩動態河川縱剖面變化之空間尺度。(三)以新研發之數值模式量化分析河槽特徵流量及建置臺灣河川水力幾何型態關係，依據初步建立之臺灣河槽幾何型態特性關係，將傳統沖積河槽河制理論延伸應用於混合河川水力幾何特性及其河槽演變分析，評估混合河川最適河槽寬度。及(四)依據濁水溪案例河川之研究經驗及成果，建立混合河川動態穩定河槽分析作業流程，可擴充應用於臺灣其它具相同地質條件及河槽型態之混合河川，解析動態穩定河槽演變特性，作為河川管理之參考。

關鍵字

動態穩定河槽 ； 軟弱岩床 ； SRH-2D ； 向源侵蝕 ； 遷急點 ； 空間尺度 ； 河相變化 ； 滿槽流量 ； 滿灘流量

並列摘要

Rivers in Taiwan are very steep and subject to a large amount of sediment transport. In the last two decades, most lower and even middle river reaches in west-central Taiwan have experienced rapid degradation and deposition caused by earthquake induced riverbed uplift, construction of in-stream structures, and mining activities. In many reaches, armor layers were washed away and underlying bedrock was exposed. A unique feature of the degradation in Taiwan is that the exposed bedrock consists primarily of soft bedrock which is even less resistant to shear and abrasive erosions. As a result, the bedrock-exposed reaches experienced degradation characteristics different from the widely known channel evolution of the alluvial streams. Then, a mixed alluvial and bedrock stream formed in the downstream of hydraulic structures such as dams and weirs. There is an urgent need in Taiwan to understand the dynamic channel evolution under the purely soft-bedrock or mixed alluvial-soft-bedrock conditions. Such an understanding would help planning and developing engineering solutions to protect the upstream infrastructures.
In this study, a systematic approach is developed to assess the dynamic stable channel under the engineering viewpoint of a mixed alluvial-soft-bedrock stream. First, both new alluvial and bedrock bank lateral erosion model as well as bedrock vertical erosion module are developed and coupled into an existing two-dimensional (2D) depth-averaged flow and sediment transport model SRH-2D. The new 2D model has the capability to simulate the combined vertical and lateral erosion for a mixed alluvial and soft-bedrock stream. Then, both empirical and numerical modeling approaches are adopted to evaluate the dynamic stable longitudinal channel profile. The temporal evolution of the longitudinal channel profile is assessed numerically with three spatial scales: The large (the entire study reach), the medium (four sub-reaches), and the local (cross-sections) scale. Next, the dynamic stable channel width is assessed through empirical formula, flow regime theory and numerical modeling within the medium spatial scale. The calculated results shows that the dynamic stable channel width keeps almost the range through different methods. Finally, the dnamic longitudinal channel profile and dynamic channel width are integrated into dynamic stable channel for the study reach. The proposed stable channel is identified through mobile-bed scenario modeling.The results shows that the proposed stable channel can reduce erosion depth significantly.
According to the developed procedure in this study, the dynamic stacble channel geometries are determined as followings: (1) After river channel incision into soft bedrock, the channel slope of the head-cutting sub-reach increases in response to the manmade cross-stream structure; soft-bedrock channel develops an average slope of 0.0058 while the downstream alluvial channel evolves to an average slope of 0.0068. The transition sub-reach is found to have its channel slope decreased. (2) Dynamic stable channel width: The dynamic stable channel width for head-cutting, bedrock-erosion, mixed-transition and downstream alluvial sub-reach are 350 meter, 420 meter, 450 meter and 550 meter, respectively. The study results can be used to select the proper locations and types of engineering stabilizing structures.
The main academic contribution of this study is to analyze the historical channel geometry and predict the future channel evolution in the mixed alluvial and soft-bedrock river through the new developed numerical model, SRH-2D. The innovations of this study include: (1) Development of a new two-dimensional numerical model to simulate both vertical and lateral erosion simultaneously in a mixed alluvial and soft-bedrock river. This newly developed model is verified and demonstrated its capability to predict the river channel evolution in a complex dynamic river. (2) The temporal evolution of the channel profile in a mixed stream is computed numerically with three spatial scales (the large, the medium and the local scales). With the medium scale analysis, the longitudinal channel profile is found to follow a predictive trend if the reach is partitioned into four distinctive sub-reaches. It is suitable for assessing the variation of longitudinal channel profile in the mixed alluvial and soft-bedrock rivers. (3)Through the new advanced numerical modeling, the relation regarding characteristic discharge and hydraulic geometry of most Taiwan rivers can be found. Then, the regime theory of alluvial channel could be extended into the mixed alluvial and soft-bedrock streams to evaluate the optimum channel width. And (4) Based on the modeling experience and results, we establish a general standard operation procedure for dynamic stable channel evaluation in Taiwan. The proposed procedure could be applied to other reaches with the same geological conditions and river setting.

並列關鍵字

dynamic stable channel ； soft bedrock ； SRH-2D ； head-cutting ； knickpoint ； spatial scales ； river morphology ； bank-full discharge ； channel-full discharge

參考文獻

1. Odgaard, A.J. and Jorge D. A. (2008). ²River meandering and channel stability², American Society of Civil Engineers, 439-459.

2. Odgaard, A.J., (2018). ²River Stability², Training Materials of Water Resources Agency, Taiwan (深槽水理特性分析應用於治理規劃之研究」教育訓練課程簡報)，經濟部水利署水利規劃試驗所(2018年12月05日，臺中)。

3. Chang, H.H. (1988). ²Fluvial Processes in River Engineering², John Wiley & Sons, New York, NY, USA.

4. Leopold, L.B., Maddock, T. (1953). ²The hydraulic geometry of stream channels and some physiographic implications², U.S. Geol. Surv. Prof. Pap. 252.

5. Fola, M.E. and Rennie,C.D. (2010). ²Downstream Hydraulic Geometry of Clay-Dominated Cohesive Bed Rivers², J. Hydraul. Eng., 136(8): 524-527.