本文利用山坡潛感分析針對可能造成邊坡破壞的潛在因子進行評估，以分析 該區域發生山崩的潛在可能性。一般研究方法可分為定性的專家評分法及定量的統計分析法，通常後者較具客觀性，因此較能被接受。雖然如此，在眾多影響因子中，由於資料無法完全收集的限制，忽略了可能造成邊坡不穩定的因子，因此定量因子的部分很難完全不考慮，且挑選的因子中，因子間的相依性仍是急待克服的問題；故本研究希望能利用岩石工程系統並配合地理資訊系統的統計及落石危害度評估系統的檢核，探討岩石工程系統的可行性，並以921集集地震引致中部橫貫公路東勢─德基段的坍方邊坡為研究案例。 主要研究成果如下：(1)災前後之預測崩塌地準確率分別為89.58﹪與84.92﹪，高潛感區分佈於東側達見砂岩及地質構造段，顯示構造造成的地質破碎及岩性本身是影響高山地區崩塌的重要因子；其次，褶皺因子是豪雨來臨時，促使邊坡崩塌的主要因素，但地震發生時，坡度則是促使邊坡崩塌的主要因素；(2)降雨對因地震而崩塌之邊坡穩定性影響極大；(3)利用RES所得之邊坡評分值RMII得知邊坡8、邊坡21、邊坡25、邊坡27及邊坡29為優先整治邊坡；(4)岩石工程系統融入山崩潛感分析的因子分級尚佳，適用於不同型式的邊坡且能共同評分比較；(5)利用現場調查之資料所得之邊坡穩定判斷，加上山崩潛感分析的檢核及輔助，可提高對邊坡穩定性的正確判斷。 Landslide susceptibility (LSS) is firstly carried out by a Geographic Information System (GIS) in this thesis to analyze the factors that may potentially account for the landslides of the Central Cross-Island Highway from Tungshih — Techi section, where severely landslide hazard had occurred after the 1999 Chi-Chi earthquake. Secondly, Rock Engineer System (RES) combined with the result of GIS analysis and Rockfall Hazard Rating System (RHRS) are used to rank the hazard potential for 58 landslide outcrops investigated after the Chi-Chi earthquake. Finally, the results of the three methods are compared and discussed. The research results reveal: (1) the result of GIS method shows that 89.6% (before Chi-Chi earthquake) and 84.9% (after Chi-Chi earthquake) of the landslides area are located within the moderately high and high susceptibility zones. Geological structure (fold) and rock formation (lithology) are the main factors controlling the landslide potential before the earthquake. Slope gradient, rock formation and fault are main factors for landslide during the earthquake in this section; (2) the result of Rock Engineer System indicates that this method is more flexible and suitable for different movement types of slope failure; (3) the slopes (outcrop number 8, 21, 25, 27, and 29) investigated in this study are most hazardous slopes according to the Rock Engineer System; (4) It’s ok in the result of Rock Engineer System, which merges landslide susceptibility’s factor grade. The method can suit to different type slopes, and to compare as well. (5) The cooperation of landslide susceptibility analysis and field investigation can improve the reliability of the result of slope stability assessment.
