南投縣仁愛鄉山崩潛感與投89縣道之穩定性分析

Landslide Susceptibility of Ren Ai Village Nantou County and Stability Analysis of County Highway 89

周宗逸(Zong-Yi Jhou)

指導教授 ： 林基源

朝陽科技大學/理工學院/營建工程系/碩士(2014年)

摘要

台灣位於菲律賓海板塊與歐亞大陸板塊交界地帶，地質年輕且造山運動劇烈，加上921集集大地震鬆動了山林的土石，特別是中部山區。近年來極端降雨出現的頻率逐漸增加，導致複合型災害發生的規模與頻率亦隨之增加。南投縣山坡地面積幅員廣大，每逢颱風豪雨，常造成慘重災情，而仁愛鄉地區之新生村、南豐村、發祥村、力行村、大同村、精英村與投89縣道沿線，常因颱風降下豪大雨，導致邊坡崩塌、滑動及土石流等嚴重災害發生。
本研究旨在建立南投縣仁愛鄉地區之山崩潛感圖，利用地理資訊系統（Geographic Information System；GIS）將高程、坡度、坡向、地質、距斷層距離、距褶皺距離、距水系距離、距道路距離與累積雨量等九項潛感因子進行分級，再與水保局2001年委託工研院所判釋之桃芝崩塌地套疊，統計出各因子中的崩壞百分比，藉由不安定指數法建立出各網格之權重值與評分，本研究將潛感區分為高、中高、中、中低、低潛感共五個等級，並繪製山崩潛感圖。最後使用誤差矩陣來判釋正確率之評估，而判釋結果：山崩正確率達74.76%、非山崩正確率達84.63%、總正確率達84.43%，顯示出檢核效果良好。
另外，本研究採用TRIGRS之邊坡穩定模式，針對投89線道路之道路中心線向外延伸500公尺之範圍進行分析，用來驗證投89線道路沿線之山崩潛感圖。最後將TRIGRS之分析結果與山崩潛感圖進行誤差矩陣之崩塌趨勢相近率判釋，結果顯示，崩塌趨勢相近率達79.58%、非崩塌趨勢相近率達98.67%、總崩塌趨勢相近率達97.37%。整體而言，本研究所建立之崩塌地潛感分析可作為未來山崩潛感分析之參考依據。

關鍵字

山崩潛感 ； 不安定指數法 ； 地理資訊系統 ； 投89線 ； TRIGRS

並列摘要

Due to locating between the Eurasian and Philippine Sea Plates and the young geology, the orogenesis events in Taiwan is abundant. Moreover, 921 Chi-Chi earthquake caused the loosed geology condition, especially in the center mountain area of Taiwan. The increasing frequency of extreme rainfall also leads to complex disaster happens more and more often. Nantou County is famous for its large slope land. Whenever there is a heavy rainfall occurred by the storm, it causes seriously disaster including landslide and debris flow. Such as Xinsheng Village, Nanfeng Village, Faxiang Village, Lixing Village, Datong Village, Jingying Village and along the Nantou County Highway 89 in Ren Ai Township.
The purpose of this study was to set up the landslide susceptibility map of Ren Ai Township in Nantou County. Using Geographic Information System (GIS) to grade the nine susceptibility factors including Elevation, Slope Angle, Aspect, Geological Situation, Distance to Fault, Distance to fold, Distance to River, Distance to Road and Accumulated Rainfall, based on overlap with the Typhoon Toraji landslide area that identified by Industrial Technology Research Institute which was authorized by Soil and Water Conservation Bureau in 2001. To calculate the landslide percentage of each factor and to acquire the weight and grade the grid by means of Instability Index Method. In this study, susceptibility can be classified into five grades: high, medium high, medium, medium low and low. And we drew the Landslide Susceptibility Map as well. At the last, a classification error matrix for accuracy assessment was then produced. The result of interpretation appeals that the Landslides accuracy is 74.76%, the correct rate of non-landslide is 84.63%, the accuracy rate is 84.43% in total.
In addition, this study use Slope Protection mode of TRIGRS model to analyze the area which was set 500 meters outward from the road center line of Nantou County Highway 89. And verify the Landslide Susceptibility Map of Nantou County Highway 89. Then, we interpreted the similarity rate of landslide tendency of the result of TRIGRS model and Landslide Susceptibility Map with classification error matrix. The result shows that the similarity rate of landslide tendency is 79.58%, the similarity rate of non-landslide tendency is 98.67%, and the similarity rate of landslide tendency is 97.37% in total. On the whole, the Landslide Susceptibility analysis in this study can be provided as references in the future.

並列關鍵字

Instability Index Method ； GIS ； Nantou County Highway 89 ； Landslide Susceptibility ； TRIGRS

參考文獻

8.方俊傑，「應用不安定指數法於坡地崩塌之潛感分析」，國立中興大學，2006年。

13.吳佐川，「台灣地區崩塌地區域特性之研究」，國立台灣大學，1993年。

15.吳柏毅，「台灣中部地區崩塌地影響因子之分析研究」，國立中興大學，2006年。

19.林伯勳、許振崑、冀樹勇，「集水區土壤厚度經驗式應用分析」，中興顧問社，2011年。

27.陳昆廷、蔡光榮、王宣惠、林欽川，「多變量不安定指數分析法應用於屏東山區道路邊坡崩塌潛感評估模式之建置研究」，中興工程季刊，第一百期，第62-72頁，2008年。