We will show why a three-dimensional model integrating crustal and mantle dynamics is required to understand Himalayan tectonics, and an example of such a model. Himalayan tectonic models were dramatically impacted by the discovery of the South Tibet fault in the mid-1980s. This structure has many characteristics consistent with normal faulting, providing a simple challenge: how to explain a major extensional fault in Earth’s most dramatic contractional orogen? Early hypotheses were two-dimensional (in cross section) and focused on the crust, with subduction treated as a boundary condition within a static geometric framework. Later work showed that the South Tibet fault was largely active as a backthrust atop a duplexing system which constructed the Himalayan crystalline core. Within this context, a vexing observation was the inconsistent duration of faulting along the South Tibet fault. It is long known that the fault ceased motion in the western Himalaya ~20 million years ago, but in the east-central Himalaya ~12 million years ago, with intermediate ages between. We show that the younging towards the east-central Himalaya is systematic, and correlative to records of south Tibetan magmatism. Two-dimensional models are inadequate to address this progression along the belt’s length. We illustrate correlation of this timing data to the history of slab underthrusting, relative rollback, and lateral migrations of slab detachment inferred from seismic tomographic data. Our model includes the along-strike younging of South Tibet fault cessation and shows how slab dynamics could control the kinematics of the orogenic wedge. Additional model features are: the asymmetric Himalayan arc curvature can be explained by different rates of slab detachment migrations from the west and east ends of the belt, and the predicted mid-Miocene culmination of surface uplift is a better fit to the South Asian monsoon intensification vs. prior models of early Miocene culmination.

The face to face double subduction zones are common in nature, with the typical example of the Manila-Philippine double subduction system. The dynamic evolution of the double subduction zones may have been significantly affected by each subduction, but the interaction between the two subducting slabs in a double subduction system remains poorly understood. Based on the Manila-Philippine double subduction system, we conducted 2D thermo-mechanical numerical modeling to study the dynamic evolution of two successive and face-to-face subducting slabs, investigating especially the interaction between the two slabs. The model results show strong interaction between the two slabs. First, the initiation of the second subduction zone leads to the termination of the first one. Second, affected by the newly formed subduction, the older subduction slab becomes steeper with a much slower trench migration/convergence rate on the surface. Third, Mantle flow and upper pates play important roles in transport interactions. Since the model results agree well with the observations, we conclude that the initiation of the Philippine subduction zone could significantly affect the evolution of the Manila subduction zone.

As part of the ongoing subduction front of the Indian crust, the W-E striking northern Himalayan belt (NHB) extends around 2900km, which is separated from the southern Tibetan Plateau in the north by the Yarlung-Zangbo suture zone. The NHB has absorbed at least 112 km N-S tectonic shorting with the onset of the continent-continent collision, a process that generated a double normal thickness of the crust (80 km) beneath the northern Himalayan orogenic belt. Previous studies have mostly focused on the mechanism responsible for the crustal thickening beneath southern Tibet, however, less attention has been paid to the processes that drove crustal thickening beneath the northern Himalayas. Debates remain regarding the extent and geometry of some main thrust systems at depth (like MHT, STD and YZS) and the exposure mechanism of Yarlungshampo gneiss dome in the NHB, as well as the specific tectonic style and mechanism of regional crustal thickening. In this study, based on the seismic data that we have collected over an 120-km-long deep reflection seismic profile, together with the north-striking 114 short period dense array data that distribute from the NHB junction to the Lhasa terrane at 92.1°E - 92.2°E, we will carry out an integrated study between structural interpretation on the deep seismic reflection profile and the receiver function. Combined with previous studies in geology and tectonic chronology, a comprehensive tectonic model of the NHB is constructed. The results present: 1) a merge of the STD at depths with the MHT beneath the Gangdese weak reflection zone of southernmost Lhasa terrane, 2) the Yarlungshampo Dome resulted from exhumation of the Indian crystalline basement during subduction, and 3) processes of antiformal stacks are responsible for crustal thickening of the Indian subducting front.

The Tibetan Plateau was formed by continental collision, and amount of north-trending rifts are observed in southern Tibetan. Whether the eastward extension is mainly considered as gravitational collapse or tectonic actives remain controversial. A distinct deep crustal structure image is the key to answering this question. Our high-resolution receiver function image crossing the Cona rift highlights the extension of the Indian lower crust and a sharp Moho offset. It demonstrate that the rifts were caused by deep tectonic activity. Miocene ultrapotassic rocks are widely distributed in southern Tibet, but the currently indian plate underplating in the collision zone conflicts with the generation of ultrapotassic magmas. So the high angle dipping of the subducted indian slab was necessary that it created the wedge for asthenospheric upwelling and melt the lithospheric mantle. Meanwhile, the weakening lithosphere without the supporting of Indian subducting plate started to collapse. We propose that the eastward extension is in response to the lateral variation of indian subducting plate. 

Lithospheric weakness is essential in subduction initiation. Spreading ridges are divergent plate boundaries which may represent lithospheric weakness and promote subduction initiation. Natural examples of ridge-inversed subduction along spreading ridges have been proposed (e.g., the Proto-South China Sea subduction). Although, the dynamic evolution of ridge-inversed subduction has been investigated by geodynamical numerical modeling previously, it remains obscure, especially the influence of the thermal state and geometries of spreading ridges on subduction initiation and dynamic evolution. We establish two-dimensional thermomechanical coupled numerical models to simulate the dynamic evolution of forced subduction along spreading ridges, and quantify the influence of four major parameters on subduction development (i.e., the spreading rate and cooling age of spreading ridges, forced convergence rate and asymmetric ridge geometry). Our model results suggest the following findings. (1) The cooling age of spreading ridges and the forced convergence rate are the most important parameters controlling ridge-inversed subduction initiation, and the thresholds of the two parameters are revealed, i.e., subduction may easily initiate with a cooling age less than ~20 Myr and a forced convergence rate lager than ~4 cm/yr. (2) The spreading rate of ridges prior to forced convergence and asymmetric ridge geometries play a secondary role in subduction development. (3) Ridge-inversed subduction of the Proto-South China Sea along the Palawan spreading ridge was proposed geologically, and our numerical modeling results support this scenario.

The N-S normal faulting systems, one of the most significant active structure in the southern Tibet Plateau, represents east-west (E-W) extension of the Tibetan plateau. Several hypotheses have been proposed to interpret N-S trending rifts. There is still controversy over its formation mechanism. Pumqu-Xainza Rift, as one of the N-S trending rifts, is restricted discussed about formation mechanism due to sparse data. In this study, high-resolution P-wave receiver functions are calculated to image the crustal structure by deploying a new high-density seismic array across the northern Pumqu-Xainza Rift. Our results show that the depth of the Moho increases from ∼60 km in the east to ∼70 km in the west under the rift. The geometry structure of the rift extends along the east-dipping decollement instead of cutting through Moho. Our observations suggest that the formation of the northern Pumqu-Xainza Rift is related to the material escaping eastward. 

The Tibetan plateau is a typical result of ongoing continent-continent collision, which provides a natural laboratory for the study of collisional orogeny. The continuous uplift of the Tibetan Plateau is a consequence of the collision between the Indian Plate and Eurasian plate, but the formation of several extensional rifts in southern Tibet is still under debate. These rifts are parallel to the direction of the main collision, which is different from the traditional compressional orogenic model. To understand the deep mechanism triggering the rifting, we chose the Yadong-Gulu rift as the research object, which is the largest rift system. 107 seismometers were deployed in a line crossing the rift. The receiver function and tomography have been applied to discover the deep tectonic structures of the Yadong-Gulu rift.

We present a high-performance, low-cost force-balance accelerometer (FBA), suitable for earthquake and structural monitoring applications. This innovative, class-A force-balance instrument has been designed, simulated, constructed and experimentally evaluated, from scratch. The proposed design is based on a double spring pendulum, with a variable capacitance displacement transducer, and a double, symmetrical coil-magnet forcer. The double coil-magnet architecture eliminates any nonlinearities encountered in single coil-magnet transducers that are typically used in several existing products in the market today. The variable capacitor central (moving) and side (stationary) plates are made of ordinary PCB pieces in order to maximize cost efficiency. The seismic mass is suspended from the frame through two parallel leaf springs, and it mainly consists of the central sensor plate and two attached coils. The two leaf springs and the mass comprise a totally symmetrical design that ensures rejection of any cross-axial motion. The two N45 magnets are firmly attached on the rigid aluminum frame of the instrument. Furthermore, the electronic control circuit is implemented on a small, high-density (4-layer) PCB and is mounted on top of the mechanical construction, while its ground plane is oriented in a way that blocks EMI generated by the coils. The instrument’s closed loop transfer function in the frequency domain, as well as all stability and frequency response issues were thoroughly investigated. The performance of the accelerometer was also evaluated through recordings of true, local seismic events (time and frequency domain analysis). The proposed FBA demonstrates an excellent performance, comparable to high cost commercial FBAs, providing an output voltage proportional to ground acceleration with a flat response from DC to over 200Hz, differential output range ±20V (40Vpp), and sensitivity 2.5V/g covering a range of ±4g. 

The 2008 Ms8.0 Wenchuan earthquake and 2013 Ms7.0 Lushan earthquake both occurred along the the Longmen Shan fault zone in eastern Tibet. It is necessary and urgent to know the variation of the stress state in the source area before these two strong earthquakes. Since the b value is inversely related to the regional tectonic stress level, and the seismic apparent stress can be viewed as the lower limit estimation of the regional tectonic stress level，our present study aimed to obtain the spatial and temporal distribution characteristics of b values, apparent stresses of small and medium events along the Longmen Shan fault zone. Along with available results of repeated seismicities (as an indication of the creeping state), we finally analyzed the variation of stress state with time before the Ms8.0 Wenchuan and the Ms7.0 Lushan earthquakes along the seismogenic fault zone. In conclusion, low b-value in the source areas were revealed before both events, and sections of decreasing of b-values were further identified. Interestingly, the observed decrease was found to disappear about 1.0~1.5 years before the major events. Additionally, within the regions of decreasing of b-values, at least one ML≥4.0 earthquake with high apparent stress occurred about one year before both events. This finding confirms that the seismogenic faults were under relatively high stress level. Meanwhile, the repeated earthquakes were all located in sections with high b-value or relatively high b-values, indicating that the stress level in the creeping section is comparatively lower. We suggest that these above characteristics can be used to evaluate the strong seismic hazard of an active fault zone. The research was jointly supported by National Key R & D Program of China (2018YFC150330501) and National Science Foundation of China (41574047).