Identification of the Slopeland Failure Types by Geometric Form Index

• Main Authors: Chia-Ling Huang, 黃佳翎
• Other Authors: Su-Chin Chen
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/51830219092516632393

碩士 === 國立中興大學 === 水土保持學系所 === 105 === The landslide and debris flow behavior are closely related to terrain conditions and both induced by rainfall. In order to understand the characteristics are of these slope hazards and to determine the difference between landslides and debris flow, it is imperative to the topographic features of slope failure. The purpose of the study is to improve the accuracy of image recognition on landslides and debris flow and to provide a reference for further analysis. This study use Taiwan and Kaoping River basin as the research area, using major disaster cases from 2006 to 2016. Slope failure brought by Typhoon Morakot in Kaoping River basin and used to analyses the topographic features of deep-seated landslides, shallow landslides and debris flow. The results show that factors may be used for identification including disaster length, width, depth, area, volume, slope and elevation. The area is the most important factor in Taiwan, while in Kaoping River basin the length is the most important factor. In terms of location and distribution, it is found that slope failure occurs on the hillside and toward the toe. Deep-seated landslides and debris flow occured in the hillside. Landslide length divided by the slope length define the scale. An increase in the slope failure area, the scale also increases, but the same is not true with slope in the range 30 - 50 degrees. In this study, the geometric patterns of slopeland disaster are presented with V= BxLxH, where V is disaster volume, B is disaster width, L is disaster length, and H is disaster depth, and these are given by B=α_1 V(β_1 ) , L=α_2 V(β_2 ) , H=α_3 V(β_3 ). The results show that the individual average of indices β_1 and β_2 of the shallow landslide is about 0.76. Debris flow is in the coefficient α_2>α_3>α_1 and the exponential β_1>β_2>β_3. For deep-seated landslide β_3 will be significantly larger than β_1 andβ_2, and when the value of β_1 and β_2 plus is about 0.55. After correction by the relationship between landslide deflection and length of the landslide in Taiwan disaster cases, the accuracy of discriminant analysis was 0.82, and the accuracy rate of Typhoon Morakot in Kaoping River Basin reached 0.95. This study suggests that the method of using geometric form index to identify disaster characteristics can effectively classify the slope failure.