Rifting and strike-slip shear in central Tibet and the geometry, age and kinematics of upper crustal extension in Tibet

The youngest deformation structures on the Tibet Plateau are about NNE-trending grabens. We first combine remote-sensing structural and geomorphological studies with structural field observations and literature seismological data to study the Muga Purou rift that stretches at c. 86°E across central Tibet and highlight a complex deformation field. ENE-striking faults are dominated by sinistral strike–slip motion; NNE-striking faults have normal kinematics and outline a right-stepping en-echelon array of grabens, also suggesting sinistral strike–slip; along NW-striking fault sets, the arrangement of grabens may indicate a dextral strike–slip component. Thus, in central Tibet, rifts comprise mostly grabens connected to strike–slip fault zones or are arranged en-echelon to accommodate sinistral wrenching; overall strain geometry is constrictional, in which NNE–SSW and subvertical shortening is balanced by WNW–ESE extension. The overwhelmingly shallow earthquakes only locally outline active faults; clusters seem to trace linkage or propagation zones of know structures. The earthquake pattern, the neotectonic mapping, and the local fault–slip analyses emphasize a distributed, heterogeneous pattern of deformation within a developing regional structure and indicate that strain concentration is weak in the uppermost crust of central Tibet. Thus, the geometry of neotectonic deformation is different from that in southern Tibet. Next, we use structural and palaeomagnetic data along the Zagaya section of southern central Tibet to outline significant block rotation and sinistral strike–slip SE of the Muga Purou rift. Our analysis supports earlier interpretations of reactivation of the Bangong–Nujiang suture as a neotectonic strike–slip belt. Then, we review the existing and provide new geochronology on the onset of neotectonic deformation in Tibet and suggest that the currently active neotectonic deformation started c. 5 Ma ago. It was preceded by c. north–south shortening and c. east–west lengthening within a regime that comprises strike–slip and low-angle normal faults; these were active at c. 18–7 Ma. The c. east-striking, sinistral Damxung shear zone and the c. NE-trending Nyainqentanghla sinistral-normal detachment allow speculations about the nature of this deformation: the ductile, low-angle detachments may be part of or connect to a mid-crustal décollement layer in which the strike–slip zones root; they may be unrelated to crustal extension. Finally, we propose a kinematic model that traces neotectonic particle flow across Tibet and speculate on the origin of structural differences in southern and central Tibet. Particles accelerate and move eastwards from western Tibet. Flow lines first diverge as the plateau is widening. At c. 92°E, the flow lines start to converge and particles accelerate; this area is characterized by the appearance of the major though-going strike–slip faults of eastern-central Tibet. The flow lines turn southeastward and converge most between the Assam–Namche Barwa and Gongha syntaxes; here the particles reach their highest velocity. The flow lines diverge south of the cord between the syntaxes. This neotectonic kinematic pattern correlates well with the decade-long velocity field derived from GPS-geodesy. The difference between the structural geometries of the rifts in central and southern Tibet may be an effect of the basal shear associated with the subduction of the Indian plate. The boundary between the nearly pure extensional province of the southern Tibet and the strike–slip and normal faulting one of central Tibet runs obliquely across the Lhasa block. Published P-wave tomographic imaging showed that the distance over which Indian lithosphere has thrust under Tibet decreases from west to east; this suggests that the distinct spatial variation in the mantle structure along the collision zone is responsible for the surface distribution of rift structures in Tibet.

青藏高原上最新的变形构造为大致呈北北东（NNE）向展布的地堑（grabens）。本研究首先将遥感构造与地貌学研究、野外构造观测以及文献中的地震学数据相结合，对位于西藏中部、约沿86°E延伸的木格普若裂谷（Muga Purou rift）开展研究，并揭示出复杂的变形场。北东东（ENE）向断层以左旋走滑运动为主；北北东向断层具有正断运动学特征，并构成右阶雁列地堑阵列，同样指示左旋走滑作用；沿北西（NW）向断裂带，地堑的展布样式则可能反映出右旋走滑分量。因此在西藏中部，裂谷多为与走滑断层带相连的地堑，或以雁列状排布以调节左旋扭动作用；整体应变几何形态呈收缩型（constrictional），即北北东-南南西（NNE-SSW）向与近垂直方向的缩短，被北西西-南东东（WNW-ESE）向的伸展所平衡。绝大多数浅源地震仅在局部勾勒出活动断层；地震簇似乎追踪了已知构造的连接或传播带。地震活动模式、新构造填图以及局部断层滑动分析均表明，发育中的区域构造内部存在变形的分散、非均匀分布模式，并显示西藏中部上地壳的应变集中作用较弱。因此，新构造变形的几何形态与藏南地区存在显著差异。

随后，我们利用藏南中部扎嘎亚（Zagaya）剖面的构造与古地磁（palaeomagnetic）数据，揭示出木格普若裂谷东南侧存在显著的地块旋转与左旋走滑作用。本研究分析支持了此前将班公湖-怒江缝合带（Bangong–Nujiang suture）重新活动为新构造走滑带的解释。

接着，我们梳理了现有关于青藏高原新构造变形起始时间的研究，并提供了新的地质年代学（geochronology）数据，认为当前活跃的新构造变形始于约5 Ma前。在此之前，约18–7 Ma期间，区域内存在走滑与低角度正断层组合下的近南北向缩短与近东西向拉伸作用。近东西向展布的左旋当雄剪切带（Damxung shear zone）以及北东向展布的念青唐古拉左旋正滑脱断层（Nyainqentanghla sinistral-normal detachment），为该类变形的性质提供了推测依据：韧性低角度滑脱层可能属于中地壳滑脱层（décollement layer）的一部分或与之相连，走滑断层即植根于此；也可能与地壳伸展作用无关。

最后，我们提出了一个运动学模型，用以刻画青藏高原新构造的质点流动过程，并对藏南与藏中构造差异的成因进行了推测。质点从西藏西部向东加速运移；随着高原的扩张，流线首先发生发散。在约92°E处，流线开始汇聚且质点加速，该区域以藏中东部主要走滑断层的出现为特征。流线转向东南方向，并在阿萨姆-南迦巴瓦（Assam–Namche Barwa）与贡嘎构造结（Gongha syntaxes）之间达到最大汇聚程度，此处质点运动速度最高。在两构造结以南，流线发生发散。该新构造运动学模式与十余年来基于GPS大地测量（GPS-geodesy）得到的速度场具有良好的相关性。

藏中与藏南裂谷构造几何形态的差异，可能源于印度板块俯冲相关的基底剪切作用。藏南近乎纯伸展构造区与藏中走滑-正断层构造区的边界斜穿拉萨地块（Lhasa block）。已发表的P波层析成像（tomographic imaging）结果显示，印度岩石圈（lithosphere）向西藏下方俯冲的距离自西向东逐渐减小，这表明碰撞带沿线地幔结构的显著空间差异，是青藏高原裂谷构造地表分布格局的主控因素。