Summary: | 博士 === 國立成功大學 === 資源工程學系 === 104 === Terrain change (landslide volume) is a major indicator in analyzing the scale of landslides. Its importance lies in assisting disaster prevention personnel to understand landslide scale and magnitude, and to estimate the possible scope of subsequent impact in the future. Two main calculating methodology of terrain change are DTM and empirical method. Even though the empirical method may have faster calculating capability, but it has restriction on applied the general empirical function to all the regions and could not offer any spatial variation of terrain change. In the first part of this work, the instability index method is applied for landslide susceptibility analysis. The result shows that major landslide triggering factor of shallow landslide is slope angle, whereas for deep seated landslide is the amount of accumulated rainfall for the case at southern Taiwan. In the second part of the work, genetic algorithm is applied to perform a smart method of calculating landslide volume automatically. The result shows that the calculation error is less than 25% and the method could offer the calculating result within a short time (within a day).
|