道路邊坡崩塌潛勢評估-以台21線、台14線為例

Evaluation of Slope Failure Potential along Route Tai-21 and Tai-14

溫健淳(Chien-Chun Wen)

指導教授 ： 陳皆儒

共同指導教授 ： 劉家男

國立暨南國際大學/科技學院/土木工程學系/碩士(2014年)

https://doi.org/10.6837/NCNU.2014.00083

摘要

本研究選擇多次受颱風豪雨重創之省道台21線(範圍由路線樁號52K+000至143K+000)及台14線(範圍由路線樁號63K+500至98K+500)道路邊坡作為崩坍潛勢探討之研究對象，探討崩塌潛勢預測與風險地圖於預警之應用。首先建立此兩路段之崩塌與未崩塌案例資料庫，考量誘發崩塌發生之8個因子（坡度、坡向、坡高、地層種類、距斷層距離、距水系距離、植被覆蓋面積百分比和颱風累積降雨量），共建立472筆道路邊坡資料以供分析。經進一步的因子獨立性檢定後選定出6個相互獨立因子作崩塌潛勢分析指標。本研究採用邏輯斯迴歸分析以及鑑別分析二種崩塌潛勢預測模型進行分析，邏輯斯迴歸分析所得之正判率為81.9%、誤判率為18.1%；而鑑別分析所得之正判率為72.5%、誤判率為27.5%，顯示二種方法對於道路邊坡崩塌判釋均有不錯的成果。所得結果進一步繪製出崩塌潛勢圖與風險分級圖可以探討其於崩塌預警之應用，此方法可根據各道路邊坡易致災特性以及預估的累積降雨量，事先評估道路邊坡的崩塌機率及其風險等級，便以掌握坡地災害高潛勢與風險較高路段，提升道路邊坡災害預警能力。

關鍵字

邏輯斯迴歸分析 ； 鑑別分析 ； 崩塌潛勢 ； 風險分級

並列摘要

Two segments of provincial routes that were prone to disaster during typhoon or heavy rainfall in Nantou County were selected for evaluation of the slope failure potential. The slopes considered in this study included those along Route Tai-21 (from Chainage 52K+000 to 143K+000) and Route Tai-14 (from Chainage 63K+500 to 98K+500). The first task is to put together a landslide inventory for the study. For each slope site, values of eight factors (including slope angle, slope aspect, slope height, geological formation, distance to a fault, distance to a river, vegetation coverage ratio, and the cumulative rainfall) were extracted and compiled. As a result, data for 472 slopes (stable and unstable) were reduced. Subsequent analysis on independence of factors using the compiled data in the database suggested that six of them are independent, and they are used for calibration of landslide potential evaluation models. Two models, logistic regression analysis and discriminant analysis, were adopted to study the landslide potential for the two routes. The accuracy rate for the logistic regression model is 81.9% and that for the discriminant analysis is 72.5%, which suggests that both approaches can obtain satisfactory results for landslide potential evaluation. With the calibrated evaluation model for landslide probability, further analysis were carried out to predict the landslide potential and to compute the risk value under certain accumulated rainfall. The computed results can then be presented in landslide potential and risk maps, respectively. These maps are useful tools for identifying high failure potential or high risk locales, which can facilitate the management and response of landslide hazards.

並列關鍵字

Logistic Regression Analysis ； Discriminant Analysis ； Landslide potential ； Risk.

參考文獻

1、Agresti, A.(2002).“ Categorical Data Analysis (2nd ed.)”, New York: John Wiley, p.p.710.

2、Carrara, A.,Crosta, G.,Frattini, P.(2008)"Comparing models of debris-flow susceptibility", Geomorphology,Vol.94,p.p.353-378.

3、Feinberg, S.(1985)“ The analysis of cross-classified categorical data (2nd ed.)”, Cambridge,MA: MIT Press, p.p.198.

4、H.J. Oh, S.Lee, W.Chotikasathien, C.H.Kim, J. H.Kwon.(2009)"Predictive landslide susceptibility mapping using spatial information in the Pechabun area of Thailand", Environ Geol, Vol.57,p.p.641-651.

5、K. T. Chau,J. E. Chan.(2005)"Regional bias of landslide data in generating susceptibility maps using logistic regression Case of Hong Kong Island", Landslides p.p.280-290.