古崩塌地演育與崩積層成因探討 –鹿場案例

Abstract

大規模崩塌有著歷時性短、毀滅性高的特性，往往造成嚴重的人員傷亡。研究老崩塌地之成因與影響可對了解未來可能發生之大規模崩塌有良好的啟發，探討古崩塌地之地形演育過程正是本研究的重點；然而因為古崩塌地發生的年代久遠、資料不齊全，此類古崩塌地地形演育上往往缺乏初始地形資料，是研究古老崩塌地通常會遭遇之難題。 本研究場址位在苗栗縣南庄鄉鹿場，本處為古崩塌地並具有高山階地特徵以及厚達100公尺的崩積層；三者之間的關係、古崩積層的成因與地形演育等疑團是本研究之重點。此研究藉由建立地質模型提出地形演育概念模型，透過Slide穩定性分析和PFC顆粒運移堆積模擬佐證力學之合理性，模擬的堆積結果與現況地形比對，不符合則反覆修正地形演育直到符合所有地質與地形的條件為止。本研究認為由西向東滑動至多只能提供巨厚崩積層總量體的三分之一，由南向北的滑動才是巨厚崩積層之主要來源。數值模擬結果可說明由南向北的滑動有機會堆積100公尺厚的崩積層，且由於倒退式滑動破壞而形成此場址所呈現之高山階地。

A large-scale landslide may occur in a sudden and result in devastating consequences with huge casualties and/or property loss. The study on the cause and effect of paleo-landslides may help to understand the principles of a landslide failure and its results. The important and relevant information about a paleolandslide is often not available or not complete; for example, the original topography of a studied paleolandslide is usually unknown. This research aims to investigate the geomorphological evolution of a paleo-landslide in Taiwan. A paleolandslide site in Luchang (in Miaoli County, Taiwan) is taken as a case study; special geological/geomorphological features including very thick (>100m) colluvium, steep scarps and alpine terrace landscape are clearly observable in this site. The objectives of this thesis study are threefold: to study (1) the origin of the very thick colluvium, (2) the geomorphological evolution of the paleolandslide, and (3) the formation of alpine terrace landscape. As a beginning, this research constructs the geological model of the site. Subsequently, the study proposes a conceptual model to explain the geomorphological evolution of the paleolandslide and the formation of the very thick colluvium. To support the reasonability of the proposed conceptual evolution model, the study adopts the software SLIDE to conduct slope stability analysis and uses the software PFC to carry out run-out simulation after landslide failure. The simulated result is compared with the current topography in this site. The conceptual model of geomorphological evolution is iteratively revised until the simulated results of the relict slope and the deposition after landslides eventually agree with the current topography. It is concluded that the deposited colluvium as a result of the west-to-east landslide can only provide one third of the actual colluvium at most. The main source of the very thick colluvium should come from the mass of the south-to-north landslide. The simulated results support that the south-to-north landslide is able to produce a 100-m colluvium; the results also indicate that the alpine terrace landscape was likely formed by retrogression landslide failure.