大地震引致之近地表變形的分析與地殼應力及強度的反衍

Analysis of Near Surface Deformation by Discrete Element Modeling and Inversion for Crustal Stress and Strength Using Stress Perturbations from Large Earthquakes

楊宜蓉(Yi-Rong Yang)

指導教授 ： 胡植慶

國立臺灣大學/理學院/地質科學研究所/博士(2013年)

https://doi.org/10.6342/NTU.2013.02692

摘要

基於地震災害的防範，近日來學者對於同震斷層擴展褶皺的研究相當有興趣，然而斷層擴展褶皺往往與地底下未出露的盲斷層錯動有直接的關係。霧峰槽溝為1999年集集大地震時，該處向東傾34度的主要斷層錯動產生地表突起的背斜型褶皺。霧峰槽溝在集集地震後於地表土壤層發展出一條主要的基底逆衝斷層、一條較短的次要逆衝斷層、一條反向的後衝斷層與土層拱起的背斜型褶皺。為了模擬鬆軟的土壤層，我們利用模擬二軸試驗，且建立了直接剪力試驗，來調整微觀參數以獲得符合土壤行為的巨觀參數，並討論顆粒鍵結類型及鍵結強度對於不同土壤類別的適用性。結果顯示，在膠結狀態下允許顆粒旋轉與相互滑動的接觸鍵結模型，可以重現軟而具塑性的土層，在盲斷層錯動時造成的地層厚度變化、複雜裂隙發展過程、褶皺前翼地層倒轉等變形特徵。模擬結果推測，霧峰地區在集集地震時，次要逆衝斷層與反向逆衝斷層先發展於地表且造成軸部的地表突起，接著主要的基底斷層才延伸出露地表，我們評估集集地震時該區的主要斷層錯動量應該為修正為6公尺。
在岩石動力學的領域裡，地殼與斷層的應力強度規模至今依舊是個基本卻未知的問題。應力規模無法被直接測量，但軸差應力則可以用大量的地震震源機制解推算出來。我們將以往的應力逆推法加以延伸，使用大量的地震震源機制資料，並且引進大地震造成的同震變形量，逆推出三維空間絕對應力場與斷層強度規模隨著深度變化的情形。我們將所建立的應力逆推方法應用在台灣1999年規模7.6集集地震事件與日本2011年規模9.0的東北地震事件。結果顯示，中台灣斷層強度規模在近底表5公里內為40–160 MPa之間，深部的地殼的斷層強度皆大於80 MPa。在考慮孔隙水壓的狀態下，估算中台灣地殼的摩擦係數在近底表12公里內皆大於0.5且逼近0.7，但在地殼深度12–20公里間其斷層強度減弱，摩擦係數小於0.5。而日本東北外海地下深度5到15公里則呈現相對較弱的斷層強度(40–90 MPa)，具有更小的摩擦係數0.2–0.5。因此東北地震規模9.0產生的能量大到足以改變日本東北外海底下背景應力狀態，造成近底表20公里內的主應力軸旋轉角度可達20度以上，故在震源附近逆斷層地震震源機制區，在震後出現許多正斷層震源機制解。相較之下，台灣地殼強度較強，呈現的性質相對日本地殼有較高的均質性。
從地質構造的角度來看，在台灣絕對背景應力逆推中，中台灣在地下深度30公里內，深度12–20公里間呈現較低的摩擦係數(μ≈0.4-0.5)，顯示在這深度間地殼強度較上下層來的弱。我們所逆推出來的背景應力反應一些台灣大地構造：歐亞板塊與菲律賓海板塊的碰撞擠壓、中央山脈高山塌陷現象、與構造脫逃作用。應力擠壓區位在花東縱谷與北港高區外圍的西部平原帶，造成現今的地震活躍帶與活動斷層帶。中央山脈在垂直剖面下呈現到三角型的應力擴張區，介於中央山脈與西部平原之間，則為剪應力區反應構造的脫逃現象。在小尺度的區域應力構造裡，西部平原壓縮區裡的台中盆地則為東西向擴張區。然而埔里盆地群則落在雪山山脈擴張區與西部平原擴張區之間的剪應力區，我們認為埔里盆地群兩側帶有左移分量的逆衝斷層錯動，造成pull-apart basin形成現今的埔里盆地群，至於盆地尺寸由北向南漸漸變小的原因，我們認為北側的雪山山脈西北東南向擴張與兩側斷層在北部有較多的左移分量，所以北邊的盆地於西北東南向有較大的拉力，造成盆地尺度的差異。

關鍵字

斷層擴展褶皺 ； 離散元素法 ； 應力反衍 ； 地殼強度 ； 集集地震 ； 日本東北地震 ； 埔里盆地群

並列摘要

Coseismic fault-propagation folds are generally associated with blind faults, which are of considerable academic interest and have recently been recognized as critical for assessing seismic hazards. The Wufeng excavation site was characterized by a major east-dipping basal thrust exhibiting a dip angle of 34° and 2 opposing vergent thrusts generated a pop-up anticlinal fold in soil cover induced by the 1999 Chi-Chi earthquake. In order to represent the soft soil behavior, we set up a direct shear simulation test to obtain valid parameters for soil. A series of 2D distinct element models possessing different bonding types and strengths was conducted to determine the deformation pattern near the surface and the evolution of the fault tip propagation. The coseismic deformation features of soft and plastic soil cover like the mutative limb thickness, complex ruptures and the overturned forelimbs in small-scale caused by the propagation of the blind fault tip were accurately predicted by contact-bond model in which grains were allowed to rotate and slip without cements breaking. Our results show that the alternative thrusts and pop-up structure developed before the main basal thrust fault ruptured through the ground surface at the Wufeng excavation site. We evaluated the slipping distance of the main fault at approximately 6 m in Wufeng during the 1999 Chi-Chi earthquake.
The magnitude of stress in the crust and the shear strength of faults are poorly known, yet fundamental quantities, in lithospheric dynamics. While stress magnitude cannot be measured directly, deviatoric stress state can be inferred indirectly from focal mechanism solutions collected before and after an earthquake. We extend a standard stress inversion method to invert for the 3D spatial distribution of absolute deviatoric stress and variation of fault strength with depth using focal mechanism solutions and coseismic stress changes produced by large earthquakes. We apply the method to the 1999 M7.6 Chi-Chi, Taiwan earthquake and the 2011 M9 Tohoku-oki, Japan earthquake. The estimated fault strength in Central Taiwan is constrained between 60 and 140 MPa in the upper 5 km of the crust and exceeds 80 MPa at greater depths. The shallow Taiwan crust above 10 km depth is relatively strong with coefficient of friction of at least 0.5 with a favored value of 0.7, assuming hydrostatic pore pressures. The inversion favors a somewhat weaker middle crust between depths of 10-20 km with coefficient of friction less than 0.5. The northern Japan forearc crust between 5 and 15 km depth appears to be weak with fault strength of 40-90 MPa, consistent with a coefficient of friction of 0.2-0.5. The Tohoku-oki coseismic stress change was large enough, relative to the ambient stress, to rotate the principal stress directions typically ~20° in the upper 20 km of the crust. The data from Japan require a heterogeneous ambient deviatoric stress field with short wavelength (~20-50 km) fluctions in primicpal stress orientations. In contrast, the ambient field of the stronger Taiwan crust is more homogeneous.
We use the focal mechanisms and stress change of the Chi-Chi earthquake to invert the 3D absolute stress field in Central Taiwan and discuss the relationship between background stress and the tectonic framework. The crust at a depth of 12–20 km shows a low friction coefficient, μ≈0.4-0.5, and is weaker within a depth interval of 0–30 km. The estimated background stress state explains the primary tectonic and regional structures of Taiwan: the plate motion of the Eurasian plate and the Philippine Sea plate, the orogenic collapse of the Central Range, the tectonic escape. The compressional stress state results from the action and reaction of compressional forces of the plate convergence located individually in the Longitudinal Valley and the Western Foothills around the hard Peikang High, causing current active faults and frequent earthquakes. Between the reaction force zone and the Central Range, the tectonic unit escapes toward the northwest, resulting in shear stress. For regional stress, we conjecture the Puli basins are a series of the pull-apart basins developed by the NE-SW extension in Hsuehshan Range adjacent to Puli basins and the local strike-slip faults on the sides. High left-slip of left-lateral faults and strong extension in the northern part of Puli basins cause these basins increase in size from the southwest to the northeast. In the south of Taichung basin, stress presents east-west extension resulted from tectonic loading.

並列關鍵字

fault-propagation fold ； distinct element method ； stress inversion ； crust strength ； Chi-Chi earthquake ； Tohoku-oki earthquake ； Puli basins ； Taichung basin

參考文獻

Allmendinger, R.W., 1998. Inverse and forward numerical modeling of trishear fault-propagation folds. Tectonics 17, 640–656.

Aoki, Y., Fukuda, J., Kato, T., 2012. Coseismic and postseismic deformation due to the 2011 Tohoku-oki earthquake, paper presented at 10th International Conference on AOGS-AGU(WPGM) Joint Assembly, Abstract SE76-A012.

Angelier J., 1979. Determination of the mean principal direction of stresses for a given fault population, Tectonophysics, 56, T17-T26.

Angelier, J., 1984. Tectonic analysis of fault slip data sets. J. Geophys. Res. 89, 5835–5848.

Angelier, J., 1990. Inversion of field data in fault tectonics to obtain the regional stress, III. A new rapid direct inversion method by analytical means. Geophys. J. Int. 103, 363–376.