摘要
约50Ma前印度板块与欧亚板块开始碰撞之后,青藏高原发生了令人瞩目的整体隆升,成为晚第三纪以来亚洲乃至全球最为重要的地质事件,并使青藏高原成为大陆岩石圈变形最为强烈的地区之一,是全球学者研究大陆动力学乃至地球动力学的焦点和热点地区。由于印度板块与欧亚板块的碰撞以及组成青藏高原各地块向东和东南的挤出运动,使位于青藏高原东边缘大凉山地块及其附近地区具有明显的高原和盆地之间的过渡带特征,地壳变形严重,地壳厚度变化剧烈,并且是重力梯度带和航磁异常明显的地区,也是(GPS)资料显示的地壳运动方向由东向东南发生转变的关键地段。本区不仅蕴藏有丰富的金属矿等矿产资源,也是我国强烈地震最为频繁的地区之一。国家973科研项目《活动地块边界带动力过程与强震预测》等以本地区为目标区,就青藏高原东边缘带的活动地块的运动性状、地块之间的接触关系及相互作用和边界断裂的深浅耦合关系,以及板内强震发生的原因进行研究。大地电磁测深法作为一种探测地球深部电性结构有效的地球物理方法,在青藏高原东边缘开展了剖面式探测研究,其中冕宁-宜宾剖面是横穿东边缘带大凉山地块的东西向展布最长的剖面,本文将以冕宁—宜宾剖面为重点,开展大地电磁(MT)研究,这也是该地区首次利用大地电磁法进行深部结构研究。为了获得该地区可靠的地壳上地幔电性结构,并研究是否存在中下地壳流动层及其地质意义,作者采用最先进的大地电磁数据处理分析技术,对观测资料进行了由定性到定量全面地分析,通过二维反演得到了沿剖面的较为详细的地壳上地幔电性结构。通过把本剖面和位于其北侧的石棉—乐山剖面和南侧的美姑—绥江剖面等的综合分析,再次给出了青藏高原东边缘带地壳流动层(“管流层”,"channel flow layer")的电磁探测证据,并结合其它地质和地球物理资料的分析,就有关地块的结构特征和接触关系、岩石圈运力学、地壳流动层和地震活动性的关系等问题进行了研究分析。1大地电磁资料分析解释技术的应用研究青藏高原东边缘带以及冕宁-宜宾剖面位于地质构造复杂、地形起伏较大和断裂(断层)非常发育的地区。在当前三维数据反演技术尚不成熟,主要利用二维技术进行数据解释的情况下,分析由于复杂的构造和结构等引起的畸变问题,认识和减少畸变对数据解释效果的影响无疑是非常有意义的。1.1阻抗张量分解技术及其应用为适应二维大地电磁资料的观测、处理和解释,在上世纪60年代提出了张量阻抗技术,替代原来的仅用于一维分析的标量技术(Hermance,1973)。同时也发现当存在三维构造和结构时,观测资料受到的畸变影响(Berdichevsky,1998),特别是在近地表存在小尺度不均匀体时,观测资料受到的局部畸变影响受到关注,并成为困扰大地电磁资料解释的难题之一直到上世纪80年代末期,阻抗张量分解技术的提出,为识别、克服或减少局部畸变的影响找到了一种新的思路(Swift(1967);Bahr(1988); Groom&Baily(1989);赵国泽(1996);王立凤(2001);晋光文(2003))。本文利用Bahr和Groom & Baily提出的阻抗张量分解技术分别对冕宁—宜宾剖面的大地电磁资料进行了分析,并与常规的Swift技术分析结果进行了对比分析,结果表明,沿剖面的大多数测点的资料受到局部畸变的影响很小,使用Swift技术进行分析是合理的。对于受到局部畸变影响的少数测点,通过利用Groom & Baily的阻抗张量分解技术进行分析,获得了反映深部区域电性结构的主轴方向(电性走向)和有关畸变参数,在资料分析和二维反演时,合理地使用区域电性主轴方向和相关畸变参数,则能够得到合理的视电阻率、阻抗相位曲线,进而反演获得真实、可靠的电性结构。1.2断裂对观测资料的影响分析为了分析不同性质断裂(带)对观测资料的影响,本文利用二维模型技术计算了具有不同宽度、不同深度和不同倾向的断裂(带)模型对视电阻率曲线等的影响,结果表明,断裂(带)两侧地层的电性差异对TM极化模式的视电阻率曲线的影响大于对TE极化模式的影响,并出现了视电阻率曲线的跳变,而断裂带宽度在一定范围内的变化对视电阻率曲线影响不大。但是,随着断裂(带)深度的增大,视电阻率受到的影响范围也增大,断裂(带)倾向的变化也会引起视电阻率曲线的变化。该研究结果,对于在断裂发育区大地电磁的观测有一定的帮助和指导作用。1.3地形影响的分析为了研究地形起伏对大地电磁资料反演的影响,本文利用二维模型反演技术,计算了地表为山谷和山峰情况下面存在低阻凸起的相对简单的模型。对于TM模式资料的反演,如果模型中不考虑地形的存在,无论对于山谷还是山峰模型,反演结果中,低阻凸起的顶界面都出现了向上的平移,并影响到深部岩层的电性结构,产生冗余的异常构造,且数据拟合差较大。而采用带地形的反演,不论是山谷还是山峰模型,都能较好地确定低阻凸起顶面深度及其下部地层的电性界面。这一研究为冕宁-宜宾剖面反演中,采用带地形的反演技术,获得较可靠的电性结构,提供了一定的理论依据。1.4充分利用大地电磁资料的信息和二维反演中采用合理的技术措施大地电磁观测可获得视电阻率、相位、二维偏离度、主轴方向和磁感应矢量等多种信息,在进行资料解释时,充分利用和分析这些信息,能够更全面地对地下的电性结构进行约束,获得更真实的地质解释结果,特别是在象冕宁—宜宾剖面所在的复杂构造地区,这种分析就更有必要。通过分析,对沿剖面的电性结构的变化程度、分段性、分层性、曲线畸变特点以及电性走向等有了第一手的定性分析结果,不仅可以为反演中数据和参数的选择提供依据和帮助,而且把它们和二维反演结果相结合,可确保关于地下电性结构解释结果的可靠性。在二维反演中,分别使用TE、TM、TE加TM三种模式的资料进行了反演,经反复对比确定使用TE模式加TM模式资料的联合反演。此外,对于反演模型网格的剖分密度、数据背景误差的设定、模型光滑度参数的选择等也进行了多次对比试验,最后选择了合理的反演参数和模型网格的剖分。本文最后得到可接受的电性结构模型,就是利用带地形的NLCG(非线性共轭梯度法)二维反演方法,对TE加TM资料联合反演得到的。2冕宁-宜宾剖面2D反演及电性结构在对冕宁-宜宾剖面大地电磁资料进行反演时,针对本剖面构造比较复杂的特点,我们采用分步骤逐步实施的措施,既保证反演迭代能够稳定的收敛,又避免陷入局部极值和产生冗余的结构。例如,反演时首先选用较大的光滑度因子,获得分辨率较低的较为光滑的反演模型,即获得真实模型的主要轮廓。然后,以此模型为初始模型,采用较小的光滑度因子重新进行反演,获得进一步的细化模型。按此原则进一步由小至大选择不同光滑度因子,继续反演,直到数据拟合达到满意的要求为止。在这个过程中,反演模型的网格也进行修正,即在构建下一步反演的初始模型时,不仅将各测点下的电阻率信息提取出来,而且对上一步得到的各测点的电阻率值在更细的网格上进行插值,得到新的电阻率模型作为初始模型,于是每一步反演中都包含前一步真实模型信息,不包含可能出现的局部极值或冗余电性构造的信息,经多次修改和反演迭代,从而得到真实的地下电性结构模型。反演得到的冕宁-宜宾剖面的电性结构表明,沿剖面可分成三个区段,自西向东分别对应康滇地轴、大凉山地块和四川盆地,它们之间的边界分别对应大凉山断裂和峨边断裂。康滇地轴包括安宁河断裂东西两侧附近的区域,西边界深入到川滇地块内的东部,东边界即大凉山断裂。康滇地轴的上地壳表现为高电阻率,其厚度由西向东逐渐变薄,并在与安宁河断裂带对应的部位出现了近垂向低阻带,上地壳高阻层之下为低阻层。大凉山地块的上地壳也为相对高阻层,但是其电阻率小于康滇地轴上地壳的电阻率,厚度也较小,在相对高阻层之下出现低阻层,并与康滇地轴中下地壳的低阻层连通。四川盆地的地壳整体表现为高阻层,不存在地壳低阻层。低阻层沿剖面的一个明显的特点是,康滇地轴和大凉山地块的的中下地壳的低阻层形态显示为向上凸起的拱形结构,拱形结构的顶点位于大凉山地块的中部。此外,大凉山断裂带也显示为具有一定宽度的近垂直的低阻带。甘洛断裂和西河—美姑断裂也分别对应低阻带,并象安宁河断裂和大凉山断裂一样,其底部尖灭于地壳低阻层的顶部。经分析认为,中下地壳的低阻层是由于部分熔融并可能含有流体引起的。3青藏高原东边缘带电性结构、动力学模型及地震活动性分析结合该项目中的其它四条剖面(康定,石棉—乐山,美姑—绥江,巧家等剖面)的电性结构进行综合分析,对青藏高原东边缘带有了较完整的认识。康滇地轴上地壳自北向南皆表现为电阻率大于几千Ωm的高阻层,厚度约30~40km,并自西向东逐渐减薄,至大凉山断裂处为20 km左右,反映了该区发育的上地壳古老变质岩和火成岩等。在高阻层之下为厚度20~25 km的低阻层,它自川滇地块向东,穿过安宁河断裂和大凉山断裂,与大凉山地块壳内的低阻层相连。低阻层之下,电阻率有所增大。大凉山地块相对高阻的上地壳电阻率小于康滇地轴,可能反映了该区发育较厚的古生代和中生代碳酸岩和蒸发岩。中下地壳的低阻层(高导层)的深度整体较小,其电阻率也是本区最小的层位,并形成上拱的形态,顶面最浅部约5 km深。再向东高导层深度逐渐增大,并显示向四川盆地深部倾俯的现象。高导层以下电阻率有所增大,一般为几十~几百Ωm。四川盆地电性结构同冕宁—宜宾剖面(MYp)和美姑—绥江剖面(MGp)一致,总体分为3层。第一层可分为上、下2个亚层,上亚层表现为高、低阻体横向相间排列的形式,下亚层为相对低阻层。第一层底界深度由西部的约8 km增加到东部约15 km左右。第二层表现为较厚的高阻层,其深度和厚度都显示向东逐渐增大的趋势,与中下地壳对应,内部不存在高导层。高阻层之下,电阻率又增加,与上地幔顶部对应。本区段的电性结构反映了四川盆地浅部较厚的沉积岩层之下存在厚度较大、坚硬的地壳。安宁河断裂带表现为—垂向低阻带,在中下地壳内该低阻带被高导层横切成上、下两部分,上部宽度约5 km,带内电阻率小于东、西两侧的电阻率,为几百欧姆米,其底界尖灭于中地壳低阻层顶部。低阻层以下为近垂直的相对简单的电性边界,其东侧的电阻率值高于西侧。在石棉—乐山(SLP)剖面,安宁河断裂和大凉山断裂、甘洛断裂交汇于一处,高导层以上断裂的深度比冕宁—宜宾(MYp)剖面的深度小。大凉山断裂带同样为近垂向的低阻带,宽度约几公里,断裂带底界与壳内低阻层相连。峨边断裂对应两侧具有较大电阻率差异的地块边界,深部有向东倾俯的趋势。基于青藏高原东边缘带存在管流层("Chanel flow layer")的假说,以及本区中下地壳高导层的几何形态和GPS速度分布,推测高导层的运动特点如下,青藏高原受到南侧印度板块的碰撞作用而向北运动,在北侧塔里木地块等的阻挡下,组成青藏高原的各子地块发生相对于华南地块发生向东的运动。青藏高原东边缘带的向东、向东南流动的地壳高导层由于受到东侧四川盆地的阻挡,运动方向发生向东南的转变,同时出现向四川盆地上方的仰冲和向深部的倾俯。该区电性结构和地震资料分析表明:地壳内低阻层导电性出现横向差异的地区往往对应地震较活跃的地区。例如,安宁河断裂带是强地震分部带,且小震密集分布。大凉山断裂带和峨边断裂带也是地震较强活动的地区。而在大凉山地块内部,历史纪录的中强地震少,小震活动也比较弱。松潘甘孜地块的高导层自地块中部沿着地块走向向南东东—南东方向运动,受到龙门山断裂带和四川盆地的正面阻挡,形成稳定的“T”字型结构,难以发生运动和变形,但却易于积累地震应力。因此,在汶川地震前,龙门山断裂存在很小的滑动速率以及很弱的地震活动。而当应力积累超过了地壳岩石破裂强度时,导致汶川特大地震的发生。
The India-Eurasia collision about 50Ma ago and the continued northward indentation of India has resulted in a conspicuous uplift of the Tibetan plateau as a whole, where the continental lithosphere is experiencing most intensive deformation. It is one of the most mysterious geological events since late Tertiary time in Asia and the world. Thus this region has become an idealized field laboratory for continental dynamics. The Sichuan-Yunnan region in the eastern margin of the Tibetan plateau is a transitional zone from highland to basin in tectonics because of the India-Eurasia collision and the eastward indentation of blocks in the Tibet plateau, which is characterized by steep gravity gradients from south to north and aeromagnetic anomaly belt, large variations of crustal thickness, strong crutsal movements, active faults and frequent earthquakes as well as metallic mineralization zones. GPS data reveal that the crustal motion turns to east and southeast in this region, which may imply an eastward flow of the lower crust material there.It is, however, not clear what is the deep background for the dynamic processes mentioned above. Hence to probe the structure beneath this region by geophysical, including magnetotelluric (MT) sounding, is an urgent task. This thesis focuses on analysis of the MT data collected on the Mianning-Yibing profile across the eastern margin of the Tibetan plateau, which is the longest peofile in the eastern margin of Tibet plateau, and the first study to detect the deep structure in this area.The purpose in this project is to infer deep structure beneath the study area and study whether there is a lower-crust flow layer and its implications below this area. In this thesis, the data analysis from qualitative to quantitative with the advance MT data processing technique and 2D inversion are performed to yield fine electric structure.In conjunction with the Shimian-Leshan profile in the north and Meigu-Suijiang profile in the sourth, this thesis suggests the MT sounding evidence for the channel flow layer below the study area. Consideration of the electrical structures of other profiles in this region, a detailed discussion is made on the structural characters and contact relationships among the blocks. In combination with other data (geology, seismology, heat flow, and GPS), a kinematic model is suggested for this region and a further analysis is made on seismicity in this region.1 MT data analysis and its applicationThe eastern margin of Tibet plateau is the area with complicated structures and strongly deformed strata and development faults especially where the Mianning-Yibin profile crosses. It's important to analysis, distinguish and decrease the affect of the distoration by the complicated structures when the 2D inversion is used today and 3D inversion is not applied.1.1 Impedance tensor decomposition and its application The impedance tensor was adopted in the sixth decade of last century, and took place of the impedance scalar quantity (Hermance,1973) in order to use the 2D MT data observation, processing and interpration. The local distortion (Berdichevsky,1998) caused by 3D near-surface abnormal bodies has been a difficult issue in MT data interpretation since a long time.In the eighth decade of last decade, methods of impedance decomposition were put forward as an effective tool to tackle this problem (Swift(1967);Bahr(1988); Groom&Baily(1989); Zhao(1996); Wang (2001); Jin (2003))).This work uses the impedance decomposition methods suggested by Bahr and Groom&Baily to make a comparative analysis of measurement data along the profile and contrast to the traditional method by Swift. The result shows that the distoration caused by the local anomaly body is little for majority of sites, and the true electrical structure can be acquired through the Swift analysis method. The azimuth of the principal axis on deep electrical structure and the correlative distortion parameters can be gotten by the GB impedance decomposition method to the some distorted sites.The reasonable apparent resistivity and impedance phase can be acquired with the azimuth of the principal axis and other distortion parameters.The 3D local distoration can be avoided to get the true electrical structure and the reason about the distoration is analyzed combined with the observation data.1.2 Influence of faults on MT dataThe influence of faults with variational width, depth and tendence on the MT data is calculated in this thesis in order to anyalyze the influence of different faults with the 2D forward modeling. Forward modeling shows that lithology variation between both sides of a fault can produce major influence on apparent resistivity curves and there's a bigger leap transform along the appearent resistivity curves in the TM mode than TE mode, while the width of the fault does not seiriously. But such influence would enhance with increasing depth of the fault and cause the appearent resistivity change with the varying tendence of the fault. It would give a good help and instruction to the MT data observation in the area with development structures.1.3 Influence of topography on MT dataThe relatively simple model with a low resistivity horst below the valley and the peak model is calculated with forward methods in order to anyalyze the influence of different topography.The forward modeling on relatively simple models with valley and peak topography indicates that for 2D inversion of apparent resistivity and impedance phase, the low-resistivity body below the relief shifts upward when the inversion does not include topography, which has some effect on the underlying strata, easily to produce redundent structure and big RMS in final data fitting. When the topography is considered in NLCG inversion, the low-resistivity body in the lower part of the model with a valley and peak can exhibit good response, consistent with its real position and its lower boundary can also be well revealed. It demonstrates that the NLCG inversion with consideration of topography can overcome the influence of relief effectively and acquire relatively true electric structure in the subsurface.1.4 Information from the MT data and reasonable metoods used in 2D inversionThe 2D skewness, azimuth of the principal axis, magnetic induction vectors, apparent resistivity and impedance phase are acrquired from the MT data. It's very important to define and get the true electrical structure when these informations are analysed and adopted in the data inversion calculation especially in mountainous areas where the Mianning-Yibin profile cross. With this analysis, some characteristics of subsurface electric structure can be determined, such as segmentation, layering, complexity, electric boundaries, dimension and orientations of principal axis, which not only provide evidence for the reseasonable parameters and data selected in 2D inversion and interpretation but also can be combined in the 2D inversion to proof the true electrical structure and reseanable interprations.This work analyses the inversions on different models including TE mode,TM model and the joint inversion of the TE/TM mode, and select the joint inversion of the TE/TM mode as the final result.Further more, the last inversion parameter selected and mesh grid subdivision are determined through lots of tests on the inversion model mesh grid subdivision density,data error selected and the model smooth regularization factor.The accepted model in this thesis is acquired with the joint inversion of TE &TM mode data by the NLCG method with topography.2 Two-D inversion of Mianning-Yibing profile and analysis of electric structureA stepping maner is adopted in order to get a stable convergent result and avoid falling into a local extreme value or reduce unexpected structure during Mianning-Yibin MT profile inversions. For example, a smooth model of low resolution is firstly established using a big regularization factor, which outlines the realistic model. Then this resulted model is used as the initial one and the value of the regularization factor is lowered for the next inversion. This process is repeated till the data fitting meets the requirement. The inversion mesh grid is changed in this process too.It indicates from the electrical structure along the Mianning-Yibin profile that the section can be devided into three parts.The first is Kangdian geo-axis, the next Daliang Shan block and the third Sichuan basin from west to east.Their boundaries are Daliang Shan fault and Ebian fault respectively.The west boundary of Kangdian geo-axis locates inside the Sichuan-yunnan block, and the eastern is Daliang Shan fault. The upper crust of the Kangdian geo-axis is featured by high resistivity, at thickness thinner from west to east. There is a low resistivity layer below the upper crustal high resistivity layer. The upper crust of the Daliang Shan block is featured by high resistivity too. But the resistivity is less than Daliang Shan block's, and the thickness is thinner than Daliang Shan block's.There is a low resistivity layer connectting with the Kangdian geo-axis' below the upper crust. The crust of Sichuan basin as a whole is highly resistant and no low-resistivity layer below the upper crust. An obvious character along the profile is that the low-resistivity layer of the middle-lower crust in the Kangdian geo-axis and Daliang Shan block looks as an upward arch architecture whose top is located in the middle of Daliang Shan block.Furthmore, the Anning He fault, Ganluo fault, Xihe-Meigu fault and Daliang Shan faults exhibit nearly vertical and broad zones of low-resistivity anomalies, whose bottom interfaces disappear in the top of the high-conductive layer of the crust. The high-conductivity layer in the middle-lower crust is presumably attributed to the partial melting or salt-bearing fluids involved.3 Electric structure, dynamic model and Seismicity in the eastern margin of the Tibetan plateauFrom the Mianning-Yibing MT profile, in conjunction with other four profiles in this region (Kangding, Shimian-Leshan, Meigu-Suijiang, and Qiaojia profiles), a fairly complete image of electric structure in the eastern margin of the Tibetan plateau has been acquired. It indicates that along the Kangdian geo-axis, from north to south, the upper crust exhibits a high-resistivity layer of over thousandΩm and the thickness of about 30-40km which becomes thinner toward east. In geology this layer represents the old metamorphic rocks and volcanic rocks in upper crust, as well as the crystalline basement consisting of highly metamorphic old strata. Below this high-resistivity layer is a 20~25 km thick layer of low resistivity, which crosses the Anning He fault and Daliang Shan fault and links the high-conductivity layer in the Daliang Shan block. Downward farther, resistivity increases to some degree.The electric resistivity of the upper crust in the Daliang Shan block is lower than that in the Kangding bock, probably associated with the thick carbonate and evaporite of Paleozoic and Mesozoic times there. Beneath them is a upward convex HCL, whose top is at depth as shallow as 5km in the middle of the Daliang Shan block. Eastward farther, this HCL tends to be deeper with the largest buried depth about 30km, and seems to underthrust toward the Sichuan basin. The thickness of HCL is general at 25km, with a consistent way especially on the Shimian-Leshan and Mianning-Yibing profiles, while not so obvious on the Meigu-Suijiang profile. Beneath this HCL the resistivity increases again by tens to hundredsΩm in general.Along the Mianning-Yibing and Meigu-Suijiang profiles, the electric structure of Sichuan Basin can be divided into three layers. The first layer is further subdivided into upper and lower sublayers.The upper one exhibits an alternating high-and low-resistivity pattern in lateral direction.The lower one is of relatively low resistivity. The bottom of the first layer is 8 km deep in the west, and increases to 15km in the east. The second layer is thick and highly resistive.The depth and thickness tends to increase toward east, corresponding to the middle and lower crust without inner high-conductivity layer. Below this high-resistivity layer, resistivity values rise up again down to the uppermost mantle. The electric structure of this area suggests that there exists thick and hard crust beneath the thick sediments in Sichuan basin.The nearly vertical Arming He fault coincides with a vertical zone of low resistivity, which is cut into upper and lower parts by a high-conductivity layer in the middle-lower crust. The upper part is 5km wide. The resistivity is several hundredΩm and smaller than that on the east and west sides.The bottom disappears in the top of a low-resistivity layer in the middle crust. Below this low-resistivity is a nearly vertical boundary of electricity, where the resistivity on the east side is higher than the west side. Along the Shimian-Leshan profile, the Anning He, Daliang Shan and Ganluo faults merge at one place whose depth is shallower then that along the Mianning-Yibing profile above the high-conductivity layer. The Daliang Shan fault also exhibits a nearly vertical zone of low resistivity which is several kilometers wide and the bottom linking with the low-resistivity layer in the crust. The Ebian fault corresponds to the boundary between two blocks with big differences of resistivity on either side, which tends to dip eastward.With reference to the hypothesis of the channel flow of lower crust in the eastern margin of the Tibetan plateau, as well as integrated data of the electric structure and GPS measurements, this thesis suggests a model of high-conductivity layer flow in crust in the study region which is described as follows. Due to the effect of the India-Eurasia collision in the south, the Tibetan plateau moves toward north. Because of the impeding of the Tarim block in the north, movements of massifs in the plateau turn to east with respect to South China. In the eastern margin of the plateau, the motion direction of the high-conductivity layer in crust changes from eastward to southeastward because of the obstruction of the Sichuan basin. As well as the HCL plunge down into the beneath of the Sichuan basin.Data analysis suggests that the areas with lateral variations of conductivity of low-resistivity layers in crust are usually active in seismicity in this region. For example, Major events as well as dense small quakes are distributed along Anning He fault. Around the Daliang Shan fault, small earthquakes exhibit a concentrating pattern. The Ebain fault is also an electric boundary.While within the Daliang Shan block, a few medium-sized quakes are documented in history and small events are less with respect to the boundary zones on its both sides.In the Songpan-Ganzi block, the high-conductivity layer moves in SSE-SE direction, crossing the 300km-long Longmen Shan fault zone in a normal direction and is hampered by the Sichuan basin. Thus a stable "T" shaped structure is formed at the central portion of the Longmen Shan which makes motion and deformation difficult, and stress easy to build up but not easy to release. Consequently, before the Wenchuan M8 event of 2008, the Longmen Shan fault zone was characterized by a low slip rate and weak seismicity. Once the long-term accumulation of stress exceeded the rupture strength of crustal rock beneath the Longmen Shan, sudden fractures and slips occurred to produce the Wenchuan quake in 2008.