A Preliminary Study of Land Classification in High-Risk Areas─A Case Study in Tseng-Wen Reservoir Watershed

碩士 === 國立成功大學 === 都市計劃學系 === 102 === Single disaster event will causing directly preliminary damage, with the dif-ferent environment conditions, the preliminary disaster often induce secondary disaster, and even induce multiple disaster and turn into the compound disasters. Some research also pointe...

Full description

Bibliographic Details
Main Authors: Ssu-YiFang, 方思怡
Other Authors: Hsueh-Sheng Chang
Format: Others
Language:zh-TW
Published: 2014
Online Access:http://ndltd.ncl.edu.tw/handle/84h56d
Description
Summary:碩士 === 國立成功大學 === 都市計劃學系 === 102 === Single disaster event will causing directly preliminary damage, with the dif-ferent environment conditions, the preliminary disaster often induce secondary disaster, and even induce multiple disaster and turn into the compound disasters. Some research also pointed out that the damage and losses caused by compound disasters, which may have geometrically multiplier effect but not just the simple addition effect. However, the estimation of disaster-potential in the past are mostly based on one single disaster, but not based on compound disaster. Therefore, the results of disaster-potential estimation may be underestimated. This study is fo-cused on disaster-potential that estimated by compound disaster, which referred to secondary sediment disaster causing by typhoon Morakot in 2009 and in Tseng-Wen catchment as a case study. By using the data of landslide potential which from “Landslide Susceptibility Zonation in Slope land of Urban Ar-ea ”(Central Geological Survey, 2011) as the base, and, citing concepts and prin-ciples from relevant research to set hypotheses and risk matrix combination rules. Therefore to assume compound disaster have three features which are spa-tial-spread, damage-cumulative, and disaster events would happened sequential or simultaneous. With the location data which disaster happened as validation, comparing the original susceptibility and susceptibility estimated by study hy-potheses. Assuming that area of high susceptibility and disasters have occurred as correct judgment, conversely, area of low susceptibility and disasters have not occurred as error judgment. And comparing the correct and error judgment hit rate as hypothesis validation. The results showed that the original correct hit rate was 53.05%; and error hit rate was 21.36%; while the hypothesis correct hit rate was 72.29%; error hit rate was 15.81%. Therefore, the result shows that the hypotheses is true, which means that compound disaster do have features of spatial-spread, damage-cumulative, and disaster events would happened sequential or simulta-neous. Therefore with using the features of compound disasters effects as disaster potential estimation, can distinguish area that whether affected by compound dis-asters or not.