台灣西南部地殼變形與地震活動相關性研究; Interseismic crustal deformation and seismic activity in Southwestern Taiwan

NCUIR > College of Earth Sciences > Graduate Institute of Geophysics > Electronic Thesis & Dissertation > Item 987654321/60655

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/60655

Title:	台灣西南部地殼變形與地震活動相關性研究;Interseismic crustal deformation and seismic activity in Southwestern Taiwan
Authors:	蔡旻倩;Tsai,Min-Chien
Contributors:	地球科學學系
Keywords:	地殼變形;台灣西南部;全球衛星定位系統;間震期;地震活動;九芎坑斷層;大地應變率;時間序列分析;雜訊分析;crustal deformation;GPS;southwestern Taiwan;interseicmic;modeling;seiscmic acivity;strain rate;Global Positioning System;time series analysis
Date:	2013-07-26
Issue Date:	2013-08-22 11:43:24 (UTC+8)
Publisher:	國立中央大學
Abstract:	高精度全球衛星定位系統（Global Positioning System, GPS）大地測量方法可用於監測地殼變動或觀測斷層活動，進而探討研究區域之地體構造特性。台灣西南部一直被認為是強烈地震災害的高危險區，本研究整合1994~2011年之地震目錄與1993~2012年GPS觀測資料，對台灣西南部進行地震背景活動度分析與間震期地殼變形模擬，以期了解該區之構造活動特性與地震潛能。考慮GPS時間序列中可能隱含之空間相關與時間相關誤差，本文中所使用的間震期之GPS觀測資料經時間濾波、共同誤差移除、時間序列分析、雜訊分析後，於東西、南北及垂直方向之精度分別為2.3 mm、1.7 mm及4.1 mm。結果指出，台灣地區之空間共同誤差於東西、南北與垂直方向分別為1.7 mm、1.5 mm與5.4 mm，而頻譜指數（spectral index, "κ" ）於三分量分別為-0.71，-0.73，及-0.76，表示台灣地區之時間序列雜訊模式為「全頻等幅雜波＋閃變雜波」。透過估計台灣西南部地區之地震活動度指標，有助於評估該區之地震潛能。使用最小平方法與最大可能估計法估計台灣西南部地區之最小完整地震規模（Mc）為1.5。根據1994～2011共17年之地震目錄，推求台灣西南部之地震活動指標，a、b值分別為2.68與0.72，相較於前人研究結果活動度皆屬偏低。估算2010與2011兩年相對於背景地震之活動度(Z值)，結果發現九芎坑斷層、六甲斷層、左鎮斷層、新化斷層與旗山斷層都呈現較低活動度，暗示這些地區有較高的地震潛能。使用Wu et al. （2007）之三維速度構造重新地震定位結果，ERH、ERZ、與RMS分別從0.71、3.19、0.25降低到0.33、1.82、0.18，定位精度相對提高很多。而地震發生的力學行為，主要在於調整區域應力，因此透過研究區域內的震源機制解，可以了解該地區的應力分布。估算該區震矩張量總和，結果指出台灣西南部之應力型態主要與板塊構造運動相關。嘉南地區雖受集集地震影響於1999～2005年間有些微應力形貌的改變，但於2006年後已緩慢回復為震前狀態。由2004～2012年經時間序列分析求得之速度場以許雅儒（2004）及 Hsu et al.（2009）之新方法推估台灣西南部之地表應變，結果指出嘉南地區之最大壓應變率落在前緣逆衝斷層帶，約-0.75～-0.9 μstrain/yr，最大主應變軸方向幾乎垂直逆衝斷層走向，呈西北西—東南東方向。高屏一帶最大縮短率落在左鎮斷層、旗山斷層、與小岡山斷層間的三角區域，為-1.4～-1.55 μstrain/yr。剪切應變方面最大值為0.8～0.9 μstrain/yr，為一平行旗山斷層東北—西南方向的帶狀區塊，推測與構造逃脫有關。潮州斷層東側區域（即中央山脈南段）則受到伸張應變，伸長軸方向為東—西方向。因地表之應變率和速度之梯度、地下之斷層幾何密切相關，台灣西南部之高應變率可能有部份原因是來自斷層幾何型貌之改變，而部分則來自無震滑移（aseismic slip）的影響。台灣西南部之間震期地殼運動速度場，在嘉南地區與高屏地區呈現不同的變形特性。在嘉南地區，水平速度從中央山脈區域的33～44 mm/yr向西遞減，在嘉義平原一帶約為0～5 mm/yr，除玉山附近受集集地震影響方向略微偏北，其餘皆為東西方向。跨過觸口斷層與九芎坑—木屐寮—六甲斷層斷層系統，速度場有15～20 mm/yr的梯度變化。自新化斷層以南段處開始加大並呈逆時鐘方向偏轉，在旗山斷層北段與潮州斷層北段中間區域為50～55 mm/yr，且於屏東平原沿海一帶由西向南約有30度方向的逆時針偏轉。本文針對嘉南地區與整個台灣西南部地區使用不同模式，模擬間震期之地殼變形行為。於嘉南地區於使用地塊模式、深埋錯位模式、二維斷層模式等不同模型模擬，結果指出該區存在一近乎水平之滑脫面，深度為8～13公里，向上延伸之逆衝斷層目前皆為鎖定狀態。觸口斷層系統位置接近九芎坑斷層而無法解析其滑移量，嘉義斷層則為未出露地表之盲斷層。其中九芎坑—木屐寮—六甲斷層斷層系統為一鎖定深度8～13公里，向東傾斜23～25°之逆斷層，目前呈現幾乎完全鎖定的狀態，斷層面上最大滑移速率為38 mm/yr，平均滑移速率為27 mm/yr。使用BLOCKS軟體採用多地塊模式模擬台灣西南部間震期之地殼變形，模型結果可解釋90％以上之地表觀測資料。其中新化斷層為右移斷層，跨斷層約有10 mm/yr之速度不連續，顯示斷層目前有潛移（creeping）活動。潮州斷層與梅山斷層目前皆處於活動度較低的狀態，其中梅山斷層為高角度右移斷層，面上滑移量約2～3 mm/yr。潮州斷層為一高角度之左移斷層，跨斷層無明顯速度變化，模型結果面上滑移量為6～8 mm/yr。左鎮斷層因缺乏地表近站，斷層模型參數較不靈敏，有明顯左移分量約12 mm/yr。旗山斷層為一向東傾斜55°～60°具走向滑移分量之逆斷層，斷層兩側之運動形式不同。東側存在一不同於嘉南地區之滑脫面，深度約8～10公里，斷層面上滑移量約25 mm/yr。斷層西側受到旗山斷層逆衝的擠壓，且左鎮斷層之左移運動與新化斷層之右移運動加速了此區地塊向西南逃脫。鳳山轉型斷層帶對高屏地區的地殼變形量或應變累積影響不大，只影響了部份測站的轉向，且該區地震活動並不活躍，應非主要孕震構造。因此旗山斷層東側大部分的速度改變應來自於斷層或褶皺沿基底滑脫面移動造成，構造逃脫僅發生於近海處有厚層沈積物且質地較軟的區域。 BLOCKS模擬嘉南地區間震期之地塊模式與台灣西南部的多地塊模式模擬結果都指出九芎坑─木屐寮─六甲斷層系統為23°～25∘向東傾斜之逆衝斷層，且具有最高的斷層滑移量約28～31 mm/yr及彈性滑移率約25～26 mm/yr，與地塊模式、深埋錯位模式、二維斷層模式等結果相近，為台灣西南部地區中最具地震潛能的斷層。由滑移速度估得之100年週期最大規模可達7.3，300年週期規模可達7.6，若整條斷層錯動造成之災害不容小覷。考慮高屏地區有活躍之地殼活動，相對於嘉南地區卻鮮少發生地震，可能是因為本區分布大量泥岩，塑性變形的褶皺取代大量斷層，大部分都發育成沿滑脫面向上分支之盲斷層構造型態，且鬆軟地層也較不易孕震。 The high precision Global Positioning System (GPS) geodetic survey technique provides an efficient tool to study active tectonics and geodynamics. The southwestern Taiwan is an active tectonic area, approximately half of the 80 mm/yr plate convergence rate is accommodated on the fold and thrust belt in the area. We use the GPS derived 1993~2012 interseismic velocity field and 1994-2011 earthquake data to study the crustal deformation and seismic potential in southwestern Taiwan. The observed GPS time series can be described by some model parameters such as linear rate, annual and semi-annual periodic motions, coseismic offsets, postseismic rate change, and exponential decay after earthquakes. Stacking of power spectral densities from continuous GPS data in Taiwan, we found the slopes of spectra (spectral index) are -0.71, -0.72, and -0.76 for the E, N, U components, respectively. It indicates the errors of continuous GPS data can be described as a combination of white noise and flicker noise. A more realistic noise model can give a better estimate of full covariance matrix and model parameters in GPS position time series. The common errors are removed by stacking data from 26 continuous GPS sites with data period more than 10 years. By removing the common errors, the precision of GPS data has been further improved to 2.3 mm, 1.7 mm, and 4.1 mm in the E, N, U components, respectively. By estimating the seismic activity indicators, we can assess the seismic potential in southwestern Taiwan. The minimum magnitude of completeness (Mc) of this area is 1.5 from the least squares estimation and maximum likelihood method. Based on a 17-year earthquake catalog from 1994 to 2011, we derive the earthquake activity indicators, a, b values to be 2.68 and 0.72, respectively in southwestern Taiwan. It reveals relatively lower seismic activity than previous studies. The 2010-2011 earthquake activity relative to the background (Z value) indicates a lower seismicity near the Jiuchiunken Fault, Liuchia Fault, Zuozhen Fault, Shinhwua Fault and Chishan Fault, which may imply a higher seismic potential in those areas. The earthquake locating precision, ERH, ERZ and RMS are significantly reduced from 0.71, 3.19, and 0.25 to 0.33, 1.82, and 0.18 by earthquake events relocation using Wu et al., (2007)’s 3D velocity model . The results of summation of the moment tensor focal mechanisms from 1994 to 2010 indicate the stress pattern of the area is mainly associated with plate motion. The stress pattern of in the Chiayi-Tainan area was slightly changed by the impact of 1999 Chi-Chi earthquake during the period from 1999 to 2005. However, it seems to be slowly recovering to the state before earthquake after 2006. Based on the 2004~2012 interseismic velocity field derived from GPS time series analysis, we use a new method (Hsu, 2004 Hsu et al., 2009) to estimate the crustal strain rate in southwestern Taiwan. The maximum dilatation rates of about -0.75～-0.9 μstrain/yr in the direction WNW-ESE are found at the frontal thrust faults belt in the Chiayi-Tainan area. In the Kaoshiung-Pingtung area, there is a shorting rate of about -1.4～-1.55 μstrain/yr at the triangle region bounded by Zuozhen Fault, Chishan Fault and Siaogangshan Fault. The maximum shear strain rate of about 0.8～0.9 μstrain/yr is detected along a NE-SW trending linear zone, almost parallel to the strike of Chishan Fault. It is presumably related to the tectonic escape. Extension in the E-W direction is observed to the east of Chaochou Fault (i.e., the southern section of the Central Range). The strain rate is strongly correlating with the variations of GPS velocity gradient and the subsurface fault geometry. Thus the high strain rate in southwestern Taiwan could be caused by the change of fault geometry or partly due to aseismicslip. The intersesmic velocity field of southwestern Taiwan reveals very different characteristics in the Chiayi-Tainan area and Kaoshung-Pingtung area. With respect to Paisha (S01R) of Penghu, the horizontal velocities in the Chiayi-Tainan area are mostly in the westward direction and show remarkably decreasing from the west to the east. In the Central Range, the horizontal velocities are 33~44 mm/yr and decrease to 0~5 mm/yr near the coast. There is a velocity gradient of about 15~20 mm/yr across the Jiuchiunken－Muchiliao－Liuchia fault system (JMLF). To the south of Shinhwua Fault, the velocities increase southward with directions become southwesterly. It reaches 50~55 mm/yr in the Kaoshiung-Pingtung area and deflects about 30° counterclockwise in the direction near the Kaoshiung-Pingtung coast. This implies the complexity of tectonic structures in southwestern Taiwan. To invert for fault geometries and slip rates from GPS velocities in the Chiayi-Tainan area, we consider three different approaches: (1) a block model, (2) a buried dislocation model, and (3) a two-dimensional fault model. The basis of model fault geometry is the thin-skinned model that proposed a décollement underneath the fold-thrust belt in western Taiwan. The modeling results of three different models all indicate the existence of a nearly horizontal décollement with depth of about 8~13 km and long-term slip rate of 42 mm/yr. All the thrust faults extend from the décollement are almost fully locked. The JMLF is a thrust fault with a locking depth of about 8~13 km and dips 23°～25°E. The average and maximum fault slip rates of JMLF are 27 mm/yr and 38 mm/yr, respectively. The Chiayi Fault is a blind thrust fault with the fault tip at 1 km depth and the dip of 20° to the east. It is difficult to resolve the slip rate and fault geometry of Chukou Fault because it is too close to the JMLF. The BLOCKS software developed by Meade and Loveless (2009) of Harvard University is utilized to study the interseismic deformation of southwestern Taiwan with multi-blocks. The multi-blocks model can well resolve and explain more than 90% of the interseismic GPS data. It is found that Shinhwua Fault is a right-lateral strike-slip fault with creeping rate of 10 mm/yr. The Meishan Fault is also a right-lateral strike-slip fault with slip rate of only 2~3 mm/yr. The Chaochou Fault is a left-lateral strike-slip fault with small slip rate of about 6~8 mm/yr. Although the fault geometry of Zuozhen Fault is not well resolved, we can still obtain a left-lateral fault slip rate of about 12 mm/yr. The Chishan Fault is a thrust fault with minor strike-slip component, dips 55°～60°E. The results of multi-block models indicate a different tectonic motion on both sides of the Chishan Fault. In the eastern side, there could be a décollement at depth of 8~10 km, different from that in the Chiayi-Tainan area. In contrast, the tectonic escape of western block may be enhanced due to the compression from thrust movement of Chishan Fault and lateral motion by Shinhwua Fault and Zuozhen Fault. The Fengshan transform fault zone (FTFZ) is not a major seismogenic structure in the Kaoshiung-Pingtung area, as the seismic activity is quite low and its impact on the strain accumulation is not significant. The observed velocity changes at the eastern side of Chishan Fault may come from the slip on the faults and folds extending from the basal décollement, and tectonic escape occurs only in the offshore area with thick sediments. All the BLOCKS modeling results indicate the JMLF is a thrust fault with dip angle 23°E～25°E, long-term slip rate of 28~31 mm/yr and slip deficit of 25~26 mm/yr, consistent with the results from previous modeling studies. With the highest fault slip rate in southwestern Taiwan, the JMLF can produce a major earthquake with maximum magnitude of 7.3 and 7.6 for 100 and 300 years period, respectively. The potential seismic hazard should not be overlooked. The widespread soft sediments and mudstone resulting in plastic deformation in the Kaoshiung-Pingtung area could be the main reason that there is active crustal deformation but low seismic activity in the area.
Appears in Collections:	[Graduate Institute of Geophysics] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	1830	View/Open

社群 sharing

Loading...