| 摘要: | 震波在通過不同阻抗之材質時，波傳速度的改變會體現於震測資料的走時資訊。 震測資料偏移(seismic migration)即是以此對波形資料進行重定位與回歸，所得成像為材質阻抗的在空間中的幾何分布，其有助人們判斷地下之材質岩性、構造，或是作為進一步逆推分析之初始模型以及用於偏移速度分析(Migration Velocity Analysis, MVA)等。
本研究使用的疊前繞射疊加深度偏移(PreSDSDM)具有時間偏移法之部分優點，此方法對速度資訊之誤差有較高的容忍度，且理論上亦不需針對高程進行靜態修正，也不需顧慮資料之聚排方式，適用三維測線，且輸出成果為深度域，不須再將成像結果由時間域轉深度域，是為一種具有潛力的偏移方式。
以此為出發點，本研究將對此方法進行原理分析、測試以及實際應用。所選用的震測資料分別為TAIGER計畫中收集的長支距陸上炸測與海上多頻道反射震測(MCS)資料。海上資料測線為TAIGER 2009 MGL0908-TST，資料量足夠且炸測點位相當密集；而陸上震測資料則是TAIGER 2008於南部Transect T4b測線施作之廣角反射震測測線，是一條橫跨133公里蜿蜒測線，有四個炸點與609個TEXAN單分量接收器，成像條件相對嚴苛。陸上資料所使用之速度模型參考自Kuo-Chen et al.(2012)的Local及Regional速度模型，以及Huang et al.(2014)所提出之速度模型，另有Roma (2018)以tau-p波場逆推所得之一維速度。
研究成果方面，海上資料的應用非常成功，可以觀察到Lin et al. (2012)等人所提出的部分構造，並於海面下1.7公里處附近發現明顯的海底仿擬反射(BSR)，或有甲烷水合物蘊藏於該地之可能。
陸上資料方面，受限於資料的數量及品質，炸點間距過大，淺部反射訊號難以收斂。另外西部平原區淺層沖積物造成之Multiple Diving Wave (Huang et al., 2013)，也對研究構成一定的難度。目前以挑選中遠支距初達波進行偏移之結果，可以觀察到西部平原區基盤之明顯象徵，以及炸點SP2東側疑似觸口斷層之構造，惟炸點SP3與SP4之間接收器密度不足，難以判斷中央山脈以下之構造。此外，使用Kuo-Chen et al.(2012)及Huang et al.(2014)速度模型進行偏移之結果，發現淺層成像收斂均出現”X”型圖樣，意味著淺層速度有高估的可能。
本研究所用之方法仍有許多成長空間，將來可朝使用天然震源進行偏移作為研究方向，亦或是引入三維速度模型，以三維的測線幾何作為成像考量等等。;The Pre-Stack Diffraction-Stack Depth Migration (PreSDSDM) algorithm was implemented with short-offset Marine Multichannel Seismic (MCS) data, the MGL0908-TST survey, and Long-offset active source land seismic from Transect T4b Line of TAIGER Project. Tomographic velocity models from Kuo-Chen et al. (2012), Huang et al. (2014), and Roma (2018) are also utilized in this study to derive 1D velocity for implementing PreSDSDM.
For deep structure imaging with seismic data. The imaging principle follows generalized imaging conditions for 3D depth migration. Resolving power of proposed approach depends on propagating velocity and frequency bandwidth of waveform, geometry of source-receiver pair and spatial aperture coverage of available data. Such kinds of details were discussed in this study.
Current results shows outstanding output for MGL0908-TST survey. Several detailed structural were distinguishable and supports several interpretation from other study. On the other hand, the results from active source land seismic survey from TAIGER project surfers from insufficient source and receiver coverage, crooked line, multiple diving wave (Huang et al., 2013), and low signal to noise ration. However, we’re still able to distinguish features that indicates certain structures, such as basement under Coastal Plain and Chukou Fault which is on the eastern side of shot point 2. In addition, apparent migration smiles could be found on the image migrated with different tomographic velocity model. Such fact implies that, current tomographic velocity models appear to overestimate velocity at the shallow part of subsurface.
Further improvement could be made in the future for better imaging results, such as adopting full 3D depth migration, improving steps for data conditioning, and to uses passive source for imaging. |
