碩士 / 長榮大學 / 土地管理與開發學系碩士班 / 106 / Most of the mountain regions in Taiwan are steep slope and poor geological conditions. Recently, human development coupled with the global impact of extreme weather, typhoons and heavy rains have caused the landslide disasters and leaded to human causalities and properties loss. Therefore, effective evaluation of vulnerability for landslide potential areas is one of the important tasks for disaster preventions. Previously, most of studies were concentrated on the large-scale vulnerability analysis in the scope of urban or townships. Therefore, this study explores the vulnerability of mountain residents to landslides.
Namasia District of Kaohsiung City were selected as study areas. The FORMOSAT-2 satellite images before and after three different typhoons in 2009 and 2013 were selected to interpretate for obtained the surface conditions and hazard log data. Three Clssifiers in ArcGIS Pro supervised image classification tool, Train Maximum Likelihood (ML), Train Support Vector Machine (SVM) and Train Random Trees (RT) with texture analysis were adopted to classify the satellite images. The Logistic regression method was utilized to construct the landslide probability assessment model. Based on the area under the receiver operating characteristic curve (ROC), the applicability of the model was also examined. In addition, according to the items of exposure, sensitivity, adaptability, landslide potential and village characteristics, a framework of complex vulnerability analysis was established. The vulnerability gradation map was drawn by using spatial analysis of ArcGIS platform.
Results of image classification show that the SVM with texture analysis were best. After Typhoons Morakot, Suli and Cannon, the AUC of landslide assessment evaluation results were 0.759, 0.991 and 0.967, respectively. Regardless of the accessibility of the settlement infrastructure, the results also revealed that the vulnerability of the exposed area is increased due to the proximity of the Debris flow potential areas and the fragility of the settlement will also increase. Whenever rainfall induces landslides, because of the small population in the northeast of Tangaanua, it has less impact on life and property, so its vulnerability is relatively low. Nangisaru has the highest vulnerability in sensitivity and adaptability, and the vulnerability is relatively high in the potential of landslide potential. Historical disaster data also show that Nangisaru was the most severely affected village.
| Identifer | oai:union.ndltd.org:TW/106CJU00019008 |
| Date | January 2018 |
| Creators | LIAO,WAN-WEN, 廖婉妏 |
| Contributors | CHEN, YIE-RUEY, TSAI, KUANG-JUNG, 陳怡睿, 蔡光榮 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 122 |
Page generated in 0.0103 seconds