本研究選定大漢溪流域為研究對象，收集921地震前的賀伯颱風前後期SPOT2衛星影像以及震後降雨事件前後期SPOT5及福衛2號影像來判釋山崩並作崩塌變異分析，利用地理資訊軟體建置山崩圖層，探討流域內崩塌的發生特性，包括區位分布及面積，並分析與PGA、岩性、雨量、坡度的關連性，期能了解地震對崩塌地發生趨勢的影響，以及後續的效應如何。在崩塌變異分析的過程中,將舊有崩塌地範圍視為持續存在，而將前後期衛星影像判釋之崩塌地範圍有交集的崩塌地，視為沿舊有崩塌地擴大之崩塌地，無交集部分，則歸類為新生崩塌地，兩者加總即為新增崩塌地。
研究結果顯示:(1)大漢溪流域崩塌地發生特性與趨勢明顯受到921地震之影響，但隨著時間漸進，邊坡逐漸趨於穩定 (2) 921地震提高了崩塌集中發生坡度，且於坡頂的地方持續發生崩塌。(3)山崩發育集中在砂頁岩區與板岩區，而崩塌比以板岩區最大，變質砂岩之崩塌比最小。(4) 崩塌地集中分布在PGA 50~249 gal的區域，且崩塌比隨著PGA的增加而上升，但隨距離地震時間之增長，高PGA區域的崩塌比逐漸地在下降，顯示地震效應正在減緩中。(4) 崩塌地與雨量的相關性並不好，並非降雨強度越高、累積雨量越多而崩塌越易發生，僅有馬沙颱風造成的崩塌大略有此趨勢。(5)在距離構造2km內，震前崩塌地有著離構造越近越易崩塌的特性，震後崩塌地則是剛好相反，直到泰利颱風事件，崩塌地才回復到震前的趨勢。(6)PGA主控著崩塌地的分佈，雖然其地震效應正在減弱中，但仍須有足夠的降雨強度與累積雨量，雨量的效應才會被凸顯出來。(7)以不安定指數法進行艾利颱風事件的山崩潛感分析時，無論有沒有考慮促崩因子，岩性因子均佔了最大的權重，且不安定指數值越高處，邊坡越容易發生崩塌，顯示該方法為合理的方法，在指數值4以上的區域可掌握約90%的崩塌地，而加入促崩因子分析時，高指數值的區域掌握崩塌地的能力會較為優異。

This proposed study, the catchment of Tahen river are chosen as the study area .I collect the earlier and later SPOT2 images of typhoon Herb before the Chi-Chi earthquake and the earlier and later subsequent rainfalls SPOT 5 and FORMOSAT-II images after the Chi-Chi earthquake to recognize the landslide and analyze the variations in the landslides . I used the GIS to construct the landslide layer, discussing the characteristics of the landslides in the basin including the location, area and analysis of the connection with PGA, lithology , rainfall and slope, expecting to comprehend the affection of the earthquake to the tendency of landslides occurrence and the aftermath. In the process of analysis, we viewed the former landslide as continuing existed, and viewed the overlapping landslides of the earlier and later analyzed satellite images as the expanding landslides along the former landslides. The rest regions are the new landslides. The assumption of these two types are so called the total landslides.
The study results indicate : (1) The characteristics and the tendencies of landslides in the Tahen river basin are obviously affected by the Chi-Chi earthquake but as time goes, the bankacy came to an equilibrium. (2) The Chi-Chi earthquake elevated the occurrence slope of landslides and the landslide continued at the top of bank. (3) The landslides gathered in the sedimentary rock and slate. In the slate had the biggest landslide ratio, in other hand, the metasandstone had the smallest landslide ratio. (4) The landslides located in the PGA 50~249 gal regions and the landslide ratio rise along with the PGA. The landslide ratio of high PGA regions descended along with the time from earthquake which shows the subsequent earthquake effect have diminished. (5) Around the structure in 2 km, the characteristic that nearer the structure easier to slide before the earthquake is opposite to that after the earthquake. The landslide returned to the tendency before the earthquake until the typhoon Talim came. (6) PGA controls the locations of the landslides. Though the affection of the earthquake was diminished, it still need sufficient rainfall intensity and accumulation to reveal the affection of rainfall. (7) When I used the instability index method to analyze the landslide susceptibility of typhoon Aere, the factor of lithology had the largest weight whether considering the triggering factors. Furthermore, the regions where had the higher instability index intend to have more landslides that indicated this method is feasible. The regions where the index value are higher than 4 have about 90% of the landslides. The command of the occurrence of landslides in high index value regions is more excellent when adding the triggering factors .

參考文獻
王文能、尹承遠、陳志清、李木青，九二一地震崩塌地現況與災害防治，九二一震災後中日土砂災害調查及治理研討會論文集，2000。
田坤國、李德河、呂正諭，阿里山公路邊坡震後之災害狀況與崩壞特性，921集集大地震週年學術研討會論文集，373-396頁，2000。
田坤國，由降雨資料預測阿里山公路邊坡破壞之研究，2001年公共工程學術研討會論文集，13-20頁，2001。
何春蓀，台灣地質概論台灣地質圖說明書，經濟部中央地質調查所，48-100頁，2003。
李錫堤，石門水庫集水區空中勘查，地質，第24卷，第4期，33-36頁，2005。
李錫堤，潘國樑，林銘郎，山崩調查與危險度評估山崩潛感分析之研究（3/3），經濟部中央地質調查所報告，226-227頁，2005。
周瑞燉，臺灣西部中新世五指山層之地層沈積及古地理之研究，台灣石油地質，第十一號，41-56頁，1974。
林季諍，應用主成份分析法與不安定指數於河川流域崩塌潛勢影響評估-以陳有蘭溪為例，朝陽科技大學營建工程系碩士論文，95頁，2007。
林書毅，區域性山崩穩定評估方法探討—以林口台地為例，國立中央大學應用地質研究所碩士論文，92頁，1998。
林慶偉，劉守恆，溫振宇，地震引發地質災害問題探討，2004年台灣活動斷層與地質災害研討會，1-10頁，2004。
林慶偉，謝正倫，王文能，集集地震對中部災區崩塌與土石流之影響，集集地震對中部災區崩坍與土石流之影響/臺灣之活動斷層與地震災害研討會，124-134頁，2002。
國家災害防救科技中心，石門水庫土砂災害問題分析（精簡版），12頁，2004。

國立成功大學防災研究中心，中華(福爾摩沙)衛星二號影像對地質環境與災害之動態監測及其資料庫建置(1/4)(期末報告書)，經濟部中央地質調查所報告，2005。
張子瑩，降雨與地震對形成崩塌區位之比較研究—以陳友蘭溪為例，國立台灣大學地理環境資源學所碩士論文，58頁，2002。
張政亮，張瑞津，紀宗吉，遙測與地理資訊系統應用於大甲溪流域之崩塌災害的調查與分析，地理研究 第43 期，101-122頁，2005。
許煜煌，以不安定指數法進行地震引致坡地破壞模式分析，國立台灣大學土木工程學研究所碩士論文，153頁，2002。
許輔仁，鯉魚潭水庫集水區之崩塌地潛感分佈研究，國立屏東科技大學森林系碩士論文，105頁，2002。
陳慶秋，降雨與邊坡崩塌關係之研究 :以阿里山公路邊坡為例，國立成功大學資源工程研究所，83頁，1995。
陳志豪，變質岩公路邊坡之破壞潛勢分析-以南橫公路啞口至新武段為例，國立成功大學資源工程學系碩士論文，124頁，2002。
曾志豪，降雨對阿里山公路邊坡破壞模式分析之影響研究，國立成功大學土木工程研究所碩士論文，151頁，2004。
黃臺豐，瑞里地震誘發之山崩，國立中央大學應用地質研究所碩士論文，79頁，1999。
楊介碩，高雄壽山地區岩石力學特性調查及坡面滑動觀測，國立成功大學土木工程研究所碩士論文，119頁，1989。
溫振宇，結合地震與颱風因子之山崩模式分析，國立成功大學地球科學研究所碩士論文，103頁，2005。
廖軒吾，集集地震誘發之山崩，國立中央大學地球物理研究所碩士論文，90頁，2000。
劉守恆，衛星影像於崩塌地自動分類組合之研究，國立成功大學地球科學研究所碩士論文，93頁，2002。

劉秀鳳，應用類神經網路評估台14線公路邊坡崩壞潛能之研究，朝陽科技大學營建工程系碩士論文，142頁，2005。
謝玉興，南部橫貫公路甲仙-天池段公路邊坡崩壞與降雨量關係研究，成功大學土木工程研究所碩士論文，225頁，2003。
鍾育櫻，921集集大地震前後降雨型崩塌地特徵之比較，國立台灣大學地理環境資源學所碩士論文，91頁，2005。
Crosta, G., Regionalization of rainfall thresholds, an aid to landslide hazard evaluation, Environmental Geology ,35(2),131– 145,1998.
Fuchu Dai and Chack Fan Lee, Landslides on Natural Terrain, Mountain Research and Development, 22, 1, 40-47, 2002.
Geli, L., Bard, P. Y. and Jullien, B., The effect of topography on earthquake ground motion: a review and new results, Bulletin of Seismological Society of America, 78, 42-63, 1988.
Harp, E. L. and Wilson, R. C., Shaking intensity for rock falls and slides: Evidence from 1987 Whittier Narrows and Superstition Hills Earthquake strong-motion records, Bulletin of the Seismological Society of American, 85, 6, 1739-1757,1995.
Harp, E.L., Jibson, R.W., Landslides triggered by the 1994
Northridge, California earthquake. Seismol. Soc. Am. Bull. 86, 1b, 319-332, 1996.
Johnson, K.A. and Sitar, N., Hydrologic conditions leading to
debris- flows initiation, Canadian Geotechnical Journal, 27, 789-801, 1990.
Keefer, D. K., Landslides caused by earthquakes ,Geol. Soc. Amer. Bull. 95, 406-421, 1984.

Khazai, B., Sitar, N., Evaluation of factors controlling earthquake-induces landslides caused by Chi-Chi earthquake and comparison with the Northridge and Loma Prieta events, Engineering Geology, 71, 79-95, 2003.
Rodriguez, C. E. Bommer, J. J. and Chabdler, R. J.
Earthquake-induced Landslides:1980-1997, Soil Dynamics and Earthquake Engineering, 18, 325-346, 1999.
Richard E. Giraud and Lucas M. Shaw, LANDSLIDE SUSCEPTIBILITY MAP OF UTAH, MAP 228DM UTAH GEOLOGICAL SURVEY, 2007.
S. Lee and D. G. Evangelista, Earthquake-induced landslide-susceptibility mapping using an artificial neural network, Nat. Hazards Earth Syst. Sci., 6, 687–695, 2006.
Tibaldi, A., Ferrari, L., and Pasquare, G., Landslides triggered by earthquakes and their relations with faults and mountain slope geometry - An example from Ecuador, Geomorphology, 11, 3, 215-226, 1995.
Varnes, D. J., Slope movement types and processes, in Schuster, R. L., and Krizek, R. J., eds., Landslides-Analysis and control , National Academy of Sciences Transportation Research Board Special Report176, 12-33, 1978.
Weissel, K.W., Stark, C.P., and Hovius, N. Landslides triggered by the 1999 Mw7.6 Chi-Chi earthquake in Taiwan and their relationship to topography, IEEE, 2001.