兩萬年以來花東海岸北段的海階演化

廣布於海岸山脈東翼的全新世海階地形強烈的展現花東海岸在數千年來活躍的構造特性，南北階地的形態差異則表現出抬升機制在空間分布上有所不同。前人多利用海階沉積物的定年資料計算區域抬升速率，用以解釋南北構造活動特性的差異，然而在海階分階、側向對比、產狀描述與階地成因等地形演化歷程則受限於花東海岸強烈的侵蝕作用與旺盛的堆積作用，海階地形不明顯、露頭出露狀況不佳，導致前人較少著墨。地表的塑形與地下構造息息相關，若能夠針對富含年代資料與高度控制的海階地形進行分析、則更有機會去深入探討區域構造活動特性。因此，本研究重新探討北段花東海岸─嶺頂至成功的海階成因，透過野外工作量測海階底岩面高度、確認各地沉積層與階地層序並採集定年樣本，最後由海水面升降與構造抬升速率所構成的相對海水面變化來分析花東海階自晚更新世以來的演化歷史，同時探討由海階形態所隱含的構造意義。透過階地形態的差異，本研究將海階系統分為南北兩群（T1-T6、TI-TV），並依據長期抬升速率的不同，將研究區域分為嶺頂區（2.5 mm/yr）、新社區（4.5-4.8 mm/yr）、長濱區（6.3-6.4 mm/yr）與石雨傘區（4.5-5.0 mm/yr），分區探討海水面與構造抬升在不同時間對各區的影響。在16-8 ka絕對海水面上升快速時期，四個分區的變化相似，皆呈現相對海進狀態，並在底岩上堆積厚層海進沉積層。8 ka後各地因抬升速率與構造活動特性不同，沉積層出露高度、出露形式、海階形態與階面階崖高度分布各有不同。嶺頂區與新社區海階面窄、海階崖高、兩期海階的生成年代跨距大，海階成因與鎖定斷層造成較長的再現週期、千年尺度大地震的抬升有關；南部長濱區與石雨傘區普遍出現小階崖、階面寬的現象，與潛移斷層造成的同震與快速的間震抬升有關。雖然四個分區的海階在抬升速率以及構造活動特性與縱谷斷層對應良好，但透過歷史地震事件與古地震紀錄所推估的瑞穗斷層與池上斷層地震再現週期以及同震垂直變形量卻與海岸地區的海階階崖高度存在差異。為了解決此問題，本研究利用Okada（1985）的彈性半空間模型進行模擬，認為嶺頂斷層若存在週期為千年尺度的地震，當地震規模到達Mw=7.4，則可以在海岸生成2-3公尺的海階崖，符合北段海階所觀察到的大階崖現象。池上斷層若要在海岸地區形成實際量測值為1.4公尺的小階崖，則地震規模須達到Mw=7.2，對潛移斷層較不可能，南段海階的抬升、小階崖的生成主要還是透過頻繁的小地震事件、配合快速的間震潛移抬升與侵蝕作用才能夠形成如此高度的海階崖。

關鍵字

海階地形；花東海岸；環境演化；相對海水面變化；構造抬升；縱谷斷層

並列摘要

Distinct staircase topography of marine terraces along the eastern flank of the Coastal Range obviously denotes the active tectonic uplift of the Huatung coast in Holocene. Differences of marine terrace occurrences between the northern and the southern Huatung coast indicate that uplift mechanisms of the terrace system vary from place to place. Previous researches put much efforts on calculating long-term uplift rates in order to scrutinize how tectonic structures deform the coastline, but the environmental evolution and how marine terraces were formed remained equivocal due to the cover of alluvial and fluvial deposits on top of the terrace surface. Since that formation of marine tarraces are strongly influenced by underlying structures, clarification of how Holocene marine terraces were formed increases the understanding of related fault systems. Field excursions across coastal and river profiles from Linding to Chengkung are conducted in this research, trying to discover record of terrace evolution. By measuring height of paleo wave-cut platforms, recognizing sediment occurrences and collecting samples for 14C dating, we reconstruct how paleo sea level change and tectonic upift affected terrace formation which imply the tectonic signature of the Huatung marine terraces. By comparing differences of terrace occurrences, including width of terrace surface and height of terrace riser, we classify marine terraces into two groups, T1-T6 for terraces in the north (Lindind to Zhangyuan area), and TI-TV for those in the south (Baxian cave to Chengkung area). Four tectonic sub-regions are identified with different long-term uplift rates, naming Lingding, Shinshe, Changbin and Shitiping region, with uplift rates ranging 2.5, 4.5-4.8, 6.3-6.4, and 4.5-5.0 mm/yr, respectively. As sea level rose quickly in 16-8 ka, all these sub-regions undergone relative transgression, accumulating thick marine deposits on the bedrock. As sea level reached its culmination at 7 ka and persisted stable since then, tectonic uplift became the dominant forcing for shaping the coast, turning the environment into relative regression. Forming with dissimilar uplift rates, marine terraces of four sub-regions occurred in different types. In Lingding and Shinshe region, marines terraces usually consist of narrow terrace surfaces and high terrace risers, suggesting that terraces were uplifted by locked faults with big coseismic uplift, possibly earthquakes of millenial time scale. In Changbin and Shitiping region, terraces are composed of wide terrace surfaces and low terrace risers, indicating deformation of creeping faults with small coseismic and fast interseismic uplift. Although the spatial distribution of four sub-regions well correlates with the segmentation of the Longitudinal Valley Fault, vertical displacements of paleo-earthquakes recorded in the Huatung Valley didn’t show much evidence of coastline uplifting events. Through the elastic half-space modelling proposed by Okada(1985), we find out that once the earthquake magnitude (Mw) reaches 7.4, the coseismic uplift of Linding and Ruisui fault could attain 2-3 m around coastal area, which matches heights of terrace risers in the north. For terraces in the south, as proposed by results of modelling, the 1.4 m terrace riser should be uplifted by a Mw 7.2 event, a magnitude that is hard to reach for Chihshang fault, an active creeping faults. Terraces in the south should be developed by high erosion rate accompanied with frequent small coseismic uplift and fast steady interseismic uplift.