大安溪峽谷河川地形變遷之研究

Abstract

大安溪峽谷位在苗栗卓蘭與臺中東勢之間，為1999年九二一地震後構造抬升與河流回春作用下切產生的峽谷地形。峽谷最晚於2007年形成，長約1公里，寬約500公尺。由於河川系統的變遷相當複雜，過去對於短時間的構造抬升與河川侵蝕作用之了解十分有限，本區又有環境監測、管理的必要，因此本研究整理了氣象、水文資料，並透過分析航空照片、現地觀察與河道邊坡量測等方法，了解大安溪峽谷的河川地形變遷特性，並建立河道邊坡的變遷模式。 研究區河道兩側的邊坡坡度介於15.7至78.5度間，主流下蝕深度約10-15公尺。就長期變遷而言，地震後的構造抬升先使河川流路受阻、寬度從大約40公尺增加至100公尺以上；2003年後受到下蝕、側蝕與搬運作用影響，快速產生峽谷地形，河道寬度也減少至大約10到30公尺之間。就短期變遷而言，由於2011年7月至2012年5月的研究期間缺乏大規模降雨、流量事件，因此邊坡變遷規模並不大。邊坡頂部因為直接受到雨水侵蝕，且岩性多屬於泥岩與頁岩，因此侵蝕量較大，研究期間的最大平均侵蝕量可達0.74公尺；坡腳若位處於河水直接攻擊的位置，平均侵蝕量最多可達到0.39公尺，但若非位在河水直接攻擊的位置，則變化量沒有明顯突出；邊坡中段變化量差異大，侵蝕與堆積作用出現的位置大致受到坡度影響，而坡度可能間接受到岩性、岩層傾角所影響。邊坡後退的機制為，受到河水侵蝕與雨水沖蝕作用影響，砂、頁岩互層產生的差異侵蝕作用，使邊坡的砂岩層常突出於上、下方的頁岩層之外，最後因為失去支撐而隨重力作用崩落，連帶使得上、下方的頁岩層更容易崩落。 本研究最後根據現地觀察與測量分析結果，建立4種河道邊坡的變遷模式。不過，水力作用與岩石強度的計算、水面下地形的測量、地形變遷影響因子的權重分析，都有待後續研究進行更深入的探討。

The position of Daan Gorge is between Zhuolan in Miaoli and Dongshi in Taichung, and it formed because of the tectonic uplifting of Chi-Chi Earthquake in 1999 and the erosion process of rejuvenation. This gorge formed before 2007, the length and width was about 1 km and 500 m respectively. Because the changes of river systems were very complex, and the understanding of tectonic uplifting and river erosion processes in short-term was limited, moreover, the environmental monitoring and managements of this area were necessary, so this research collected the meteorological and hydrological data first, then analyzed the characteristics of fluvial morphological changes of Daan Gorge by aerial photos, in-situ observations and channel cross-profile measurements. Finally, this research established the models of channel morphological changing of the channel slopes. The slopes of channel sides were between 15.7 and 78.5 degree, and the depth of the main channel was about 10 to 15 m. As far as the long-term changes, this research figured out that the widths of channels were increased from about 40 m to more than 100 m first, because of the obstructing of tectonic uplifting; after 2003, it formed gorges quickly because of the process of deepening, widening and transportation, and the widths of channels were decreased to about 10 to 30 m. As far as the short-term changes, because there were no large rainfall and discharge events between the research period from July, 2011 to May, 2012, so the magnitudes of morphological changes were not large. On the top of the slopes, because they eroded directly by rainfall, and the rock types were mainly mudstone and shale, so the retreated ranges were larger, the largest average retreat range of the slopes were 0.74 m in the research period. At the bottom of the slopes, if they were in positions where the river eroded directly, the largest average retreat range would reached 0.39 m, but if they were not in positions where the river eroded directly, the changes were not large. In the middle parts of the slopes, the differences of the changes were large, and the positions of erosion and deposition processes were mainly controlled by slopes, but slopes may influenced by rock types and dip angles of rock formations indirectly. The mechanism of slope retreats was that after the river and rainfall eroding, the sandstone layers were often protrude the shale layers by differential erosion processes, therefore, the sandstone layers finally collapsed by gravity because of lacking support beneath, making the shale layers on the top and at the bottom of sandstone layers collapse easier. This research finally established four types of channel slope changing models. However, the calculations of hydrological processes, the measurements of rock strengths and landforms under water level, and the weighed analyses of the impact factors of fluvial morphological changes all needed more analyses by further researches.