降雨誘發坡地二次崩塌之研究

Study of Rainfall-induced Secondary Landslide

呂宜靜(Yi-Ching Lu)

指導教授 ： 陳怡睿

長榮大學/管理學院/土地管理與開發學系(所)/碩士(2016年)

摘要

台灣地形陡峭、地質脆弱且位處亞熱帶地區，氣候變化多端，夏季時常有颱風侵襲，帶來集中性的豪雨，因此容易發生大量的山崩、地滑、土石流等土砂災害，加上舊有崩塌地可能再次發生崩塌，造成更嚴重的衝擊。因此，若能有效評估降雨造成之坡地二次崩塌，勢必能減少災害所帶來的傷害。
本研究以台灣南部之南化水庫集水區部分區域作為研究範圍，採用2009年至2013年期間，共九期降雨事件(含4次颱風及5期降雨事件)侵襲前後之福衛二號衛星影像，透過影像判釋技術，取得地表災害資料，再運用地理資訊系統及數值高程模型，萃取研究範圍之山脊與水系。研究中，藉由雨量資料之蒐集與推估，探討降雨、坡地擾動與坡地崩塌之關係，並分析崩塌空間分佈之特性。
研究結果顯示，研究區14幅不同時期衛星影像之判釋結果，一致性係數Kappa指標平均達0.77，具有中高程度的精確度。研究區各期降雨後之裸露地數量及裸露面積皆比降雨前增加，且崩塌地分佈多偏向山脊發育，而中大型崩塌地大多亦會朝溪流底端延伸。此外，不論雨量多寡，當研究區坡地擾動程度大於1.5時，開始誘發崩塌之當坡地擾動程度大於2.5時，則會有較大型之崩塌地發育相對於較大之雨量，較小降雨量時，需有較大之坡地擾動，才會發生坡地崩塌。
研究結果亦發現，各期降雨事件侵襲後，二次崩塌點位之數量皆比新崩塌點位數量多，且二次崩塌點位之崩塌規模，亦大於新崩塌之點位；較大崩塌規模之二次崩塌點位亦多偏向山脊發展。當研究範圍坡地擾動程度越大時，降雨誘發之二次崩塌面積有越大的趨勢，且當坡地擾動程度達1.6左右時，二次崩塌點位之崩塌規模會增大，再者，降雨誘發之二次崩塌點位大致上位於坡度 20˚~35˚之間。

關鍵字

坡地崩塌 ； 二次崩塌 ； 衛星影像判釋 ； 地理資訊系統

並列摘要

Taiwan is located in subtropical regions with changing climate. In the summer, typhoons often bring concentrated heavy rain, and then cause a lot of sediment disaster, such as landslides, slippery, and debris flow due to Taiwan’s steep terrain and fragile geology. Furthermore, secondary landslide could happen in existing landslide sites, causing more serious impact. Therefore, if we can effectively assess the rainfall-induced secondary landslide, it is bound to reduce the damage caused by the disaster.
The study area in this research is the Nanhua Dam watershed in the southern Taiwan. The FORMOSAT satellite images before and after the years 2009-2013 (including 4 typhoons and 5 rainfall events) were acquired and used. The satellite images were interpreted to obtain surface information and hazard log data. Coupled with digital elevation model, the spatial analysis module in the geographic information system was employed to extract the ridges and rivers in study area. This study explored the relationship among rainfall, slope disturbance, and landslide, and characteristics of landslide distribution.
The results of image classification for 14 satellite images show that the average value of coefficient of agreement is 0.77 at medium-high level. The results show that number and area of bare land increase after rainfall, most of the rainfall-induced secondary landslide sites are near the ridge top, and most of the medium-large landslides extend toward the low end of the stream. Furthermore, for every rainfall event, landslides are induced when degree of slope disturbance is greater than 1.5. For greater rainfall event, larger scale landslides start to induce when degree of slope disturbance is greater than 2.5. For smaller rainfall event, landslides are induced for large slope disturbance.
The results also show that the number and scale of secondary landslide are more than that of new landslide. Most of larger-scale secondary landslide sites are near the ridge top. The greater the slope land disturbance, the more likely it is that the scale of secondary landslide become greater. The scale of secondary landslide increases when degree of slope disturbance reaches 1.6. Rainfall-induced secondary landslides are substantially located in slope from 20˚ to 35˚.

並列關鍵字

landslide ； secondary landslide ； satellite image interpretation ； geographic information system

參考文獻

8.林洧全(2012) ，衛星影像判釋技術應用於山崩潛勢分析及風險評估模式建置之研究，長榮大學土地管理與開發研究所碩士論文，台南。

41.盧惠生、林壯沛、林介龍、黃良鑫、王秋嫻(2011)，台灣中部坡地不同土地利用及降雨強度土壤入滲，坡地防災學報，10(1)，10-26。

32.廖俊瑋、葉春國、詹仕堅、劉一新、廖學誠(2012)，太麻里溪集水區極端水文事件的頻率分析，工程環境會刊，28，89-103。

15.陳永森(2010)，極端氣候影響下對臺灣環境規劃與災害視覺之省思-以八八水災為例，工程環境會刊，25，33-44。

43.謝正倫、陳俞旭(2010)，二次災害傳遞過程與災害重建之新構想，中華防災學刊，2(2)，109-116。