新城斷層位於新竹市東南方，經過新竹科學園區南緣，是一具危險性及爭議性的活動斷層。中央地質調查所於2002年對該斷層做了一系列的調查，結果認為新城斷層應提升為第一類活動斷層，對竹科發展影響甚鉅。本研究以淺層反射震測法，針對斷層位置及其鄰近區域，進行詳盡的地下構造研究，探討新城斷層之全面特性，提供作為防災參考。 本研究結果認為新城斷層並未大規模逆衝出露於地表，斷層兩側地層均為平緩的楊梅層，並無卓蘭層出露。地表所見之地層破裂僅為淺層現象，並非來自深部，該斷層基本上屬於深埋地下之盲斷層，為典型薄皮逆衝理論之「斷層彎曲褶皺」。新城斷層具有如下特性： (1) 新城斷層以竹科為界，分為東北段及西南段。東北段有淺部滑脫褶皺產生，內夾小斷層破裂切穿地層。西南段發生地層撓曲拱起，以新城村作用最大，向兩端逐漸減弱。 (2) 東北段長度約3公里，因受到新竹斷層形成之側向斜坡及青草湖背斜匯入的影響，引起淺部地層較為激烈的變化，產生近地表地層之滑脫拱起及錯動，但錯動並未滲入到深部，僅為表皮現象。新城斷層在最北端，可能位新竹斷層阻擋而消失。 (3) 西南段長度約12公里，廣泛見到小撓曲突起發生，從深部到地表，可能引起若干散佈較廣的地表破裂，造成局部之淺地層變動，不過規模不大。斷層往南可能進入頭份地區，但作用力逐漸減弱消失。 本研究中所有剖面並未出現「卓蘭層逆衝至楊梅層之上」的大規模逆衝現象，斷層兩側之地層均為楊梅層，卓蘭層則埋在400公尺深度以下。新城斷層明顯地係為地下深處之盲斷層，對近地表地層之作用雖有，但程度範圍都不大。在地調所分類上，新城斷層是否一定要升級為第一類活動斷層，似乎可以再商榷。 Recently abundant evidence from geomorphologic and geologic studies has been proposed to identify and to ‘endorse’ an active fault (the Hsincheng fault) which passes through a computer-manufacturing area (the Hsinchu Science Park, HSP) in northern Taiwan. Great public concern has arisen as a consequence. The first substantive proof of the Hsincheng fault was provided by the Chinese Petroleum Corporation (CPC) in 1970’s, based on seismic reflection profiles and borehole data. This fault is no doubt a ‘thrust’ buried below 2 km deep. However, the layers above it seem smoothly and slightly warping, probably due to the ‘thin-skinned’ kind of detachment movement within this deeper ‘blind thrust’. In this paper, we use shallow seismic reflection lines, some co-siting with the CPC lines, to examine the detailed near-surface structure where the fault is expected. A total of 9 seismic lines across the fault, each about 1 km long and 1.5 km deep, are collected. Among them, three sections at the northern end show fault anomalies: one has an obvious detachment bulge with internal breakages and the other two have fault displacements occurring at shallow depths. However, other sections just show a small buckle vertically presenting in the structure without accompanied breakages. The structural layers are found laid flat without significant offsets at the expected places of the fault. Based on the seismic images, we suggest to lessen the Hsincheng fault as a ‘first-class’ active fault. The fault may have dislocations at deeper places, but the shallow structures are not much affected. These shallow structure variations are restricted to the sallow 400 m depth, which may represent the breakages inside the near-surface thin-skinned layer, which may not have ‘deep’ roots.
