新社地區河階地形與地下構造關係之研究

A Study of the Relations between the Terraces and Subsurface Structure in the Hsinshe Area

葛滿龍(Ger Mang-Long)

指導教授 ： 張瑞津 ； 王乾盈

國立臺灣師範大學/文學院/地理學系/博士(2005年)

摘要

1999年9月21日的集集大地震，觸發車籠埔斷層的活動，並在新社地區、豐原、石岡及東勢一帶造成非常嚴重的傷亡。該地震最大的特色有二：一是震央在集集，但最大的地表破裂卻在北方40公里外的新社階地上（地表）；二是以往文獻記載的車籠埔斷層是南北向的斷層，但斷層卻在豐原附近突然東轉，越過大甲溪終止於卓蘭、內灣一帶，咸認其地下必有特殊的構造意義（地下構造），因而引發學界對該地區地表與地下構造的熱烈討論。本研究結合地形資料、地球物理的折射震測法與反射法探討新社階地分布與其地下構造的形貌，其中值得一提是首度嘗試以折射震測法劃分階地分布、反射震測法計算其上升速率與由震測剖面顯示異常的地質構造，進而解釋其階地與地下構造的關係，並獲得以下幾點結論：
1）根據階面的震波速度、比高與崖高，新社階地的高位河階面可劃分成1個LH面及5個LT面。
2）由本文繪出大甲溪－東勢界線（TT-boundary）位置，其與車籠埔斷層東轉後之路徑一致。推測控制車籠埔斷層東轉的應該是大甲溪－東勢界線。
3）公荖坪斷層是盲斷層，推測其可能形成機制是受到西側的車籠埔斷層向西及向北逆衝而上時所伴隨產生的副斷層；同時該斷層亦受到北側TT-boundary之阻擋，以致斷層上盤地層因錯動擠壓而撓曲變形。
4）在原先提出的新社斷層(即新社線型)，反射震測剖面均未發現斷層錯動的現象，有的只是傾斜的地層，是為頭嵙山向斜之西翼。並據以推測該向斜可能的形成原因是受到車籠埔斷層的逆衝活動所致，將頭嵙山古沖積扇之西端予以抬升，彎曲成頭嵙山向斜之西翼，隨即因抬升褶皺而伴生之階地形成，後遭受古大甲溪的河流侵蝕而轉形成為河階崖。
5）推論頭嵙山向斜形成於紅土緩起伏面（LH面）發生之時，因車籠埔斷層快速的逆衝抬升，使得水井子階地在6.76萬年內升高800公尺，即11.83公釐/年之上升速率。(其中假設年代以賴光胤(2002)之熱螢光定年，約55.0±12.6ka，嚴謹保守地取上限6.76萬年)
6）整個新社階地的地下構造為頭嵙山向斜活動構造之一部分，囿於大甲溪－東勢界線（TT-boundary）的阻擋及伴隨著車籠埔斷層之活動，該構造未來亦將不斷地活動。

關鍵字

中 ； 車龍埔斷層 ； 新社斷層 ； 公荖坪斷層 ； 震測法 ； 階地

並列摘要

A Study of the Relations between the Terraces
and Subsurface Structure in the Hsinshe Area

Mang-Long Ger

Abstract
The 1999 Chi-Chi earthquake triggered the Chelungpu fault and caused serious casualties at the fault’s northern end, i.e. the Shihkang area, north of the Hsinshe terrace. This earthquake has two main characteristics: 1) its epicenter is near Chi-Chi, but the largest ruptures occurred under the Hsinshe terrace, 40 km north of the epicenter; 2) the Chelungpu fault is usually recognized as trending north-south, but during the earthquake, it turned to the east at the northern end near the Tachia River, and terminated at Cholan, north of the Ta-An River. There must exist a special underground structure to have caused such a particular fault. The present study combines topographic observation, the geophysical refraction seismic method, and the reflection seismic method to investigate the Hsinshe terrace and its relations with the earthquake faulting. It is the first attempt to use the refraction seismics to classify the terraces. In addition, the reflection seismics provides detailed underground structural pictures to shed light on the possible geologic connections between the terraces and the fault. The following conclusions are reached:
1) According to the refraction seismic wave velocities, correlatable in relative-elevation, and scarp-elevation, the Hsinshe terraces can be divided into 1 LH and 5 LT surfaces.
2) An underground boundary along the Tachia River and Tongshih (TT-boundary) is proposed, which is a structural high (a lateral ramp) to control the fault ruptures in this area.
3) The Gonlauping fault on the Hsinshe terrace is a blind fault, which could have resulted from the westward as well as northward movement of the main Chlungpu fault, and behaves as a back-thrust on its hanging wall.
4) The previously proposed Hsinshe fault may not exist. No clues of such a fault were found on the reflection seismic sections. The observed bending layers of this fault are merely the western wing of the Toukoshan syncline.
5) A new mechanism is proposed to explain the formation of the Shuijin terrace, which leads to the conclusion that the Shuijin surface was lifted by 800 m within 67,600 years, i.e. an uplifting rate 11.83 mm/year.
6) The whole Hsinshe terrace could be an ‘active’ fold. This terrace might be moved along with the excitation of the active Chelungpu fault.

並列關鍵字

Chelungpu fault ； Hsinshe fault ； Gonlauping fault ； seismic methods ； Terrace