| 摘要: | 本研究利用1991年至2007年底台灣強震網地震規模ML 4.5-5.5的資料，來推求P波及S波三維衰減模型以避免震源的複雜效應。所使用的資料區間包括1999年集集地震之餘震資料，因此，在中央山脈下的資料覆蓋度相當完整。在ω-2的震源模型假設及Q與頻率無關的假設下，我們固定每一地震事件的拐角頻率，以非線性逆推擬合觀測振幅頻譜1~30Hz的頻率範圍，估計出超過18,000筆的t*數值。每一t*數值皆以數值量化資料品質的好壞，以提供衰減影像逆推時適當的權重值。 逆推結果顯示，Qp、Qs衰減參數在斷層兩側，例如，車籠埔斷層、高屏斷層和潮州斷層的上下盤有明顯的數值差異。橫跨車籠埔斷層，斷層上盤的Qp值低於下盤的數值約85，而Qs值則低於下盤約110。Qp/Qs為1.2的等值線則恰好描繪出車籠埔斷層的斷層幾何形貌。在中央山脈下，一低Qp、低Qs和高Qp/Qs的衰減特徵正好對應於無震帶的位置。結合衰減及震波速度特性，我們可以排除液體對該構造區的影響，並推測無震帶主要為溫度效應所造成。而Qs衰減模型與Yamato et al., (2009)熱構造數值模型的吻合同時暗示著中央山脈下無震帶的物質可能來自因板塊碰撞擠壓而向上崛起的下部大陸地殼物質。在5-22公里的深度範圍，該低Qs在熱構造數值模型所對應的溫度為400°C-600°C。若利用Kampfmann & Berckhemer (1985)和Sato (1992,1994)的衰減值-地溫方程式估計中央山脈19公里深的地殼溫度，則中央山脈與周圍構造區的相對溫差約為75°C。此數值與Yamato et al., (2009)或是 Simoes et al., (2007)的熱構造數值型皆一致，然而其絕對溫度則高於目前台灣現有的熱構造模型約200°C。此結果暗示著適當的地殼岩石Q值與溫度估計關係式應用於本研究Q值深度範圍的地溫估計是必要的。 Qs衰減模型與台灣板塊碰撞模式及熱構造數值模型的比較明顯地反映出岩石流變學(rheology)及溫度變化的造山運動訊息。此外，也間接提供了區別岩石強弱的有用資訊。We determined the three dimensional Qp- and Qs- structure of the Taiwan orogenic belt to enhance understanding of the related tectonic and thermal structure beneath the collision zone. The inversion used t* values measured from the spectra of P- and S-waves from the dense Taiwan strong motion network for moderate size earthquakes (ML 4.5-5.5) to avoid source complexity. The time period of our data set, 1991-2007, includes the aftershock sequence of the 1999 Chi-Chi earthquake, which provides good ray coverage in the central Taiwan. Over 18,000 velocity spectra from 883 earthquakes were analyzed. A non-linear least square technique is applied to the spectra for t* determination by assuming a ω-2 source model for the frequency band of 1-30 Hz. A frequency-independent Q was assumed in this study. The corner frequency of a specific event was fixed for the corresponding stations, and a quality index was defined to assure good quality data for the inversion. The results reveal the sharp variation of Qp and Qs across the recently ruptured Chelungpu Fault, and the Kaoping and Chaochou Faults in Pingtung Plain. The Q values in the hangingwall are smaller by about 85 and 110 for Qp, and Qs, respectively, relative to the footwall. The fault geometry is distinctly delineated by the contour of Qp/Qs of 1.2, which extends to the depth of the geologically identified décollement structure. Beneath the Central Range, the low Qp, low Qs and high Qp/Qs features coincide well with the aseismic zone. The low Vp, low Qp, low Qs features within a low Vp/Vs as of about 1.65 and a high Qp/Qs as of about 1.4 suggest that the aseismic zone is related to the temperature effect rather than the fluid effect. Comparison to the recent thermo-mechanical numerical models of Taiwan shows that the aseismic low Qs zone corresponds to the exhumation of the lower crust. And the low Qs regime (high attenuation) at the depth of 5-22 km coincides with predicted temperatures of 400°C-600°C. Using the thermal equations of Kampfmann & Berckhemer (1985) and Sato (1992,1994), the temperature estimations show the same relative variation of about 75°C beneath the Central Range which is similar to the value of the thermal model from Yamato et al., (2009) or Simoes et al., (2007). However the absolute value is 200°C higher than the values of current thermal models. The higher value of this estimation suggests that the more appropriate equation for the crustal depth is needed. The Qs comparison with the major tectonic and thermal mechanical models of Taiwan reveals that the shear wave attenuation model contains comprehensive rheological and thermal information of relevance to understand mountain building processes. This technique appears particularly useful for distinguishing strong and weak crustal regions in the absence of other constraints. |
