The Research on the Estimation Model for the Landside Collapsion Probability in Shei-Pa National Park using Spational Information Technology

• Main Authors: I-Ju Lee, 李易儒
• Other Authors: Tien-Yin Chou
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/15670608186717386755

碩士 === 逢甲大學 === 土地管理所 === 93 === This study wanted to know the factors that affected the landside in Shei-Pa National Park and to build the estimation model for the landside collapsion probability by statistic method. This study collected multi-temporal images of SPOT satellite (1993, 1999, 2002, and 2004) and there were earthquake and typhoon occurred in this period. In order to improve the accuracy of image classification from river and collapse, this study calculated the threshold of slope degree from DEM . The result showed that the overall accuracy could be arrived at 90% and the kappa could be arrived at 85%. After that, this study collected RS and GIS factors and used spatial analysis technique to integrate. we used logistic regression to analyzed the data of the first of three periods and built the estimation model for the landside collapsion probability. The result showed that the accuracy of the estimation model was 76% and the import factors were those: the kinds of plant, the quality of soil, the distance from river, and few spectrum indexes. Finally, this study tested the last period data to estimate the landside collapsion and get a good result.