軟岩河道沖蝕對沖積型河川深槽化之影響

Abstract

就台灣西部主要河川而言，河道在離開山區進入西部麓山帶地區後，坡降即大幅降低，沖積型態為主的河道上土石資源大多豐富。然由於自然因素(如斷層錯動、地體抬升)或人為因素(如大規模採砂、興建跨河構造物或人為束縮河道等)的影響，可能造成沖蝕基準面改變、破壞沖淤平衡以致啟動河床之快速沖刷。當軟弱沉積岩層失去護甲層保護而直接裸露於河道上時，沖蝕作用加劇，可造成河道的快速變遷。軟岩河床沖蝕問題普遍存在於台灣西部各主要河川，已嚴重威脅河川區域的公共工程安全。相對於沖積型河川而言，岩床河道沖蝕的基礎研究仍屬相當有限，既有之研究議題多著重於長時期的河道地形演變，於軟弱沉積岩層沖蝕之行為研究更屬稀少。因此，有必要深入研究軟岩河道沖蝕對沖積型河川深槽化之影響，以了解軟岩質河床之沖蝕機制與引起河道演變之行為模式。 軟弱岩石河床的劇烈沖蝕為台灣相當獨特的現象，相較於硬岩每年公釐(mm)尺度的沖蝕速率而言，軟岩河床每年數十公分(cm)至數公尺(m)尺度的沖蝕速率所造成之地形快速變化，為進行岩床沖蝕研究的良好場址。大安溪921地震隆起河段，數年間即下切成為峽谷型態，隆起高度雖僅10公尺，但最大岩床下切深度卻超過20公尺，單一年度由遷急點變遷造成之局部河段最大之倒退距離達355公尺、最大下切深度達14公尺；八掌溪軟岩沖蝕河段由砂石開採、攔河堰、堤防及固床工的興建等人為因素所造成，短短二、三十年即造成11公里長河道由沖積型河道轉變為峽谷型深槽河道，累積下切量最大超過30公尺，最大河寬450公尺束縮至原寬度1/6，其巨大之地形變化規模造成河道上人工構造物經常性之損毀。 本研究以地體抬升受人為干擾較小的大安溪921地震隆起河段(大安溪大峽谷)，以及主要受到人為因素影響的八掌溪觸口至國道3號橋河段，做為主要之研究案例。藉由場址調查、多期地形分析及正射影像判釋的結果，剖析歷年河道的變化、河道侵蝕量及侵蝕速率、沖蝕機制及軟岩質河床的抗沖蝕特性等，據以重建前述二河段的沖蝕演繹過程，並研發軟岩質河道的沖蝕率模式，包含流功沖蝕模式、遷急點倒退速率模式、以及河岸拓寬模式。 軟岩質河床沖蝕模式係依據軟岩材料之沖蝕機制所建立，流功門檻沖蝕模式針對河床之均勻沖蝕行為，流功門檻由岩石抗沖蝕指數所決定，當水流之流功超過岩石之抗沖蝕流功門檻時，即發生岩石材料之沖蝕。遷急點倒退速率模式描述河道上由遷急點倒退所造成之沖蝕行為，當遷急點之坡降超過臨界坡降時，遷急點將發生倒退，倒退速率由水流流功與岩石材料之抗沖蝕流功所決定，臨界坡降則由歷年遷急點倒退之坡降資料所標定。河岸拓寬模式描述軟岩河岸之拓寬行為，由現地河岸邊坡破壞之現象調查歸納主要河岸破壞機制，基於邊坡破壞之力學分析，可決定穩定邊坡坡趾之容許坡趾沖蝕量，當累計坡趾沖蝕量超出容許值時邊坡即發生破壞。本研究運用上述模式於研究河段沖蝕案例，結合水深平均二維河道動床模式以進行軟岩河床沖蝕模擬，模擬結果與河道沖蝕變化趨勢相吻合。

Because of environmental changes (e.g., tectonic uplift, climate change, etc.) or human activities (e.g., hydraulic engineering, infrastructures construction, gravel mining, etc.), intense bedrock erosion and stream instability occurred in several major rivers in western Taiwan. Most of these rivers flow across the geological province of the western foothills which are composed of weak sedimentary rocks. Once the armor layer overlying the weak rock disappeared, the channel bedrock erosion progressed rapidly and caused enormous changes in fluvial landscape. The consequences of weak bedrock erosion have endangered the infrastructures across the river and along the river bank. The knowledge state of the erosion activities of a fluvial river is relatively matured. On the contrary, the fundamental study on the erosive behavior of a bedrock river is still in lack for evaluating the bedrock erosion processes. Therefore, the objective of this dissertation is to study the influence of soft rock erosion on the channelization of fluvial rivers. In contrast to typical bedrock erosion with a low erosion rate, the erosion rate of the bedrock river in Taiwan is exceptionally high. In several cases, the annual erosion rate could reach tens centimeters to a few meters in a weak bedrock river; the high erosion rate could result in significant changes in fluvial geomorphology. On the other hand, the rapid landscape changes provide an extraordinary opportunity to investigate the bedrock erosion processes within a manageable period of time. Two study reaches were taken in the dissertation. For the first one in the Taan River, a reach was 10-meters uplifted because of the 1999 Chi-Chi earthquake; afterwards, the reach incised 20 meters and turned into a gorge-type channel within a decade. The scale of the erosion due to knickpoint migration was 14 meters in depth and 355 meters in the horizontal direction (knickpoint retreat). For the second study reach in the Bachang River, an 11-km-long reach had been affected by human activity, such as engineering works and gravel mining. The channels transformed from an alluvial-type channel to a gorge-type channel within just a few decades. The maximum accumulative depth of the channel incision was up to 30 m approximately, with a meter-scale average annual incision rate. The maximum width of the river channel decreased to approximately one-sixth of its original width (448 m). Such an intense landscape change hence damaged various hydraulic facilities along the river bank and across the river. In this dissertation, these two reaches were chosen as the study reaches among the major rivers in western Taiwan. Chronological aerial photographs and digital terrain models were used to analyze the changes of the fluvial morphology include the channel planform, longitudinal profiles, channel incision, channel width, and channel cross sections of the study reaches. Based on the results of the channel-change analyses and field investigations, the channelization processes of geomorphology evolutions from the originally fluvial reaches were reconstructed. Moreover, three erosion-rate models for bedrock were proposed; these models include (1) an incision rate model, (2) a knickpoint migration rate model and (3) a river widening rate model. These erosion-rate models were applied along with a 2-D mobile-bed model to simulate the river channel change in the study reaches. The comparison of the field and numerical results reveals that the proposed models are capable of simulating the morphological change in a bedrock river.