由長支距震測資料比較截距時間-波線參數及速度頻譜速度分析方法

Abstract

本研究主要應用截距時間-波線參數( Tau-p )方法來分析長支距反射震測資料，建立該剖面部分區段速度構造，並將此結果與傳統的速度頻譜法所獲得之速度構造作比較。由於傳統的速度頻譜法僅依水平層構造模型分析反射波之走時曲線，因而在深度增加時由於反射訊號容易減弱及垂直隔距時差減小，增加我們對深層沉積物速度計算的不確定性。Tau-p方法則同時分析了反射、折射與廣角反射之走時曲線，除了水平層外，對有速度梯度之地層也可分析，因此對於速度構造有較好的控制。本研究分析不同震波訊號所建立的速度模型差異，並討論不同區域所適合運用的方法，使淺層速度構造的建立更趨完整。再進一步應用此兩種速度分析方法，分析臺灣西南海域被動性以及活動性大陸邊緣兩種不同的地體架構下收集來之長支距震測資料，探討南中國海大陸棚、大陸斜坡以及高屏大陸坡之淺部地層速度構造。 研究結果顯示， Tau-p法及速度頻譜法所獲得的速度構造與剖面中顯示的地質構造特徵都能夠有不錯的對應。而Tau-p法相較於速度頻譜法則能提供更高的速度構造解析度，且對於分析深部地層時，Tau-p法所提供深部地層之速度構造其可信度亦較高。將Tau-p速度分析與速度頻譜法的結果作比較，發現Tau-p法所獲得的結果，在速度頻譜較為混亂、甚難判定速度點時，可以輔助速度頻譜法決定出地層速度，兩種方法的結合有助於提升速度分析的準確度。由於Tau-p法同時需要折射與廣角反射的資料，對一般多頻道反射震測資料而言必需達到有效的支距長度才能進行分析，因此Tau-p法所能應用的資料有限，但本研究中分析的為2009年TAIGER計畫利用美國研究船藍賽斯號(R/V Marcus G. Langseth)於台灣西南海域收集來的長支距多頻道反射震測資料，其支距長達六公里，已能將Tau-p法有效地運用於水深不超過900公尺(小於1.2秒雙程走時)的平緩區域(例如:大陸棚與斜坡間盆地)。 分析臺灣西南海域被動大陸邊緣與活動大陸邊緣之震測資料顯示，水平方向上的速度變化較為平緩，大致隨著地質構造的分佈而改變其速度構造；垂直方向上的速度則隨著深度增加而變快。首先，在南中國海大陸棚的垂直速度構造分佈，於海床下雙程走時0 至 1.6秒速度由1500增加到3100 m/s。再者，於南中國海大陸坡之垂直速度構造分佈，海床下雙程走時0 至 1.4秒速度由1500增加到2500 m/s。於高屏陸坡之垂直速度變化，在海床下雙程走時0 至 1.4秒速度由1500增加到2850 m/s。最後，將兩種方法所獲得的成果結合，以間層速度為Y軸、海床下沉積物單程走時為X軸，使用線性迴歸求取速度梯度，觀察速度隨時間增加的變化趨勢，於被動性南中國海大陸棚的速度梯度最大，活動性高屏陸坡次之，被動性南中國海大陸坡則相對地最小。

This study uses the intercept time - ray parameter (Tau-p) method to build velocity structures from large-offset multichannel seismic reflection profile data, and compare them with the velocity structures derived from traditional velocity spectrum method. Velocity spectrum method analyzes only the reflection travel time curves based on a flat-layer model, and for deeper reflections where signals are weak and moveout times are small, velocity uncertainties become large. Tau-p method analyzes traveltime curves from reflection, wide-angle reflection and refraction arrivals, can also handle substrata with velocity gradients, thus could better control the accuracy of velocities derived for deeper sections. In this study, differences on velocity structures derived from different seismic arrivals are discussed, and a more completed solution for velocity analyses is proposed. Furthermore, this solution is applied to derive velocity structures in both the passive China continental margin and the active Gaoping slope to investigate their velocity characters. The velocity structures are usually consistent with the geological structures using either Tau-p analysis or velocity spectrum methods, if they are performed carefully, but the Tau-p method could provide better resolution and more accurate velocity model than that derived from velocity spectrum method for deep structures. Tau-p method could help velocity spectrum method on velocity picking when the velocity spectrum is chaotic. However, as Tau-p method analyzes the refraction and wide-angle reflection arrivals, large offset seismic data is required, thus this method has been rarely used to analyze multichannel seismic reflection profile data. In this study, the data analyzed were collected by the R/V Marcus G. Langseth during the 2009 TAIGER survey using a 6-km long streamer, and this study has demonstrated that the Tau-p method could be used to derive crustal velocity structures in the area where water depths are less than 900 m (or less than 1.2 sec two-way travel time). The velocity structures derived in this study show that velocity changes slightly laterally along the seismic profiles, while velocity increases with depth at both passive and active continental margins offshore SW Taiwan. The vertical velocity distribution at the South China Sea (SCS) continental shelf ranges from 1500 to 3100 m/s at 0-1.6 s (TWT) ; that at the SCS continental slope ranges from 1500 to 2500 m/s at 0-1.4 s (TWT); and the vertical velocity distribution at the Gaoping slope ranges from 1500 to 2850 m/s at 0-1.4 s (TWT). Analyses of the velocity profiles suggest that the velocity gradients is the highest in the SCS continental shelf area, while the SCS continental slope has the smallest velocity gradient.