Distribution of Active Faults and Lithospheric Discontinuities in the Himalayan-Tibetan Orogenic Zone Identified by Multiscale Gravity Analysis

Abstract

The lithospheric structure of the Tibetan Plateau and its adjacent area is a hot topic in geodynamic research. It is important to reveal the mechanism of crustal deformation and tectonic evolution of the study area. In this study, the techniques of wavelet multiscale decomposition and field edge detection were used to study the discontinuities of the lithosphere revealed by multilevel Bouguer gravity anomalies. Specifically, we evaluated the depth characteristics of the major active faults in the study area and identified 15 deep major faults that cut through the lithosphere. They are Chaman fault, Shyok suture zone, Altyn-Tagh fault, Karakash fault, Karakoram fault, Talas-Fergana fault, Kashgarr-Yeshgar transfer system, Rushan-Pshart suture zone, Sangri-Nacuo fault, Main Frontal thrust, Burmese fold belt, Yadong-Gulu fault, Gaoligong fault, Sagaing fault and Nujiang fault. We have also elucidated the tectonic mechanisms of two famous geodynamic phenomena in the Pamir Plateau. The first is the intense intermediate depth seismicity beneath Pamir-Hindukush. It cannot simply be described as the rupture of a subducted residual plate, which could be divided into two distinct tectonic units. One belongs to the Indian plate, the other to the Eurasian plate. Secondly, the mechanism of intense seismicity confined to the western upper crust of the Pamir Plateau could be explained as significant fragmentation of crustal material. Finally, and most importantly, we summarized the coupling mechanism between deep geodynamics and horizontal deformation as observed by modern geodetic techniques. In the upper mantle, the leading edge of the subducting Indian plate reached the SW boundary of Tarim basin and forms a closed structure in western Himalaya. Then, the Tibetan Plateau underwent a tectonic escape towards the east under the continuous compression between the Indian and Eurasian plates. During the process of tectonic escape, the role of the N–S direction normal faults in the Himalayan tectonic zone is limited.

1. Introduction

The Tibetan Plateau and its adjacent area are the primary part of the intercontinental deformation zone in the India–Eurasia collision system. It is composed of the Tibetan Plateau and the Pamir Plateau. The former is the youngest plateau in the world, while the latter exhibits geodynamic processes comparable to those of the Tibetan Plateau [1,2,3,4,5,6,7]. The Cenozoic tectonic evolutionary processes and kinematic characteristics of the Himalayan-Tibetan orogenic zone are mostly expressed by the development of complex fault systems, folds and blocks [8,9,10,11,12,13,14,15,16,17]. How these structures accommodate the convergence of the Indian and Eurasian plates is the key to understanding both the collision process between the two plates and the mechanisms of tectonic evolution [18,19,20,21]. There is long-term controversy regarding the roles of active faults in the process of tectonic evolution. Some scholars argue that active faults are very important because they separate the crust into a series of rigid blocks, and crustal deformation of the study area is mainly accommodated by activity along these major faults [22,23,24,25]. Others argue that the principal characteristics of crustal deformation in the study area are continuous and widely diffused. Active faults have a very limited role in adjusting the process of present-day crustal deformation [26,27,28,29]. Therefore, the detailed spatial distribution of active faults and structural discontinuities in the lithosphere is the key to understanding the role of active faults in the process of regional crustal deformation and the mechanisms of tectonic evolution.

At present, the main methods of deep tectonic exploration are geomagnetic, seismic wave, geothermal flow and gravity [30]. Among these, gravity has the advantages of higher efficiency, lower cost and larger observation scale. With the rapid development of satellite gravity technology, the accuracy and efficiency of deep tectonic exploration based on gravity field have greater development potential. The Earth’s gravity field is a superposition of geological gravity effects at different depths and with different properties. As the crustal lithosphere and upper mantle make up only a small part of the Earth, their gravity is bound to be strongly perturbed by the background gravity field, making them difficult to identify. Therefore, it is necessary to decompose the observed Earth’s gravity field. Wavelet multiscale analysis can decompose the gravity field data into an approximation and details of each order. The approximation corresponds to the background gravity field. Wavelet details with different power spectra could be identified as gravity signals from different depths [31,32].

Therefore, we have used the wavelet multiscale decomposition technique to distinguish the contributions of Bouguer gravity anomalies at different depths [31,33], which can effectively reveal the lateral heterogeneity of the crust at that location. The location of fracture zones within the crust and the boundaries of different geological bodies are usually accompanied by significant density mutations or density discontinuities, resulting in a greater rate of change in gravity anomalies at these locations. We have therefore calculated the total horizontal derivative (THD) of the Bouguer gravity anomaly for each layer as a numerical indicator of the spatial rate of density change and have shown the spatial distribution of active faults and structural discontinuities at different depths in the study area.

2. Method

2.1. Complete Bouguer Gravity Correction

A gravity anomaly is the difference between the measured gravity at a particular location and the theoretical gravity given by a reference Earth model for the same location. It is widely used in the study of density inhomogeneities inside the Earth.

To obtain accurate Bouguer gravity anomalies in the study area, we have used the complete Bouguer correction program package FA2BOUG (http://www.iamg.org/CGEditor/index.htm, 7 December 2021) [34] for continental areas with large topographic relief to perform the correction calculation. The initialized free-air gravity anomaly was provided by Sandwell et al. (2014) [35]. The topography or complete Bouguer correction has been historically performed in three steps: the Bouguer slab correction (Bullard A), which approximates the local topography (or bathymetry) by a slab of infinite lateral extent, with constant density and thickness equal to the elevation of the point with respect to the mean sea level; the curvature correction (Bullard B), which replaces the Bouguer slab by a spherical cap of the same thickness to a distance of 166.735 km; and the terrain correction (Bullard C), which consists of the effect of the surrounding topography above and below the elevation of the calculation point. The program performs Bullard A, B and C corrections, defining several zones depending on the horizontal distance ($R$) to the point where the Bouguer anomaly is to be calculated. FA2BOUG defines three zones: an inner zone ($R<∆x_i/2$), an intermediate zone ($∆x_i/2<R<R_i$) and a distant zone ($R_i<R<R_d$). The basic grid cell for the proximity correction calculation in the inner zone is a square area with the calculation point as the center and $∆x_i$ as the side length. $R_i$ and $R_d$ are the restricted distance for the range of the intermediate zone and distant zone, respectively.

2.2. Wavelet Decomposition on Gravity Anomalies

In order to study the tectonic features in a limited depth range such as the crustal lithosphere, it is necessary to decompose the total gravity anomaly. The wavelet multiscale decomposition method could decompose the ground total gravity anomaly into the wavelet approximations $A_K$ and the corresponding multiorder wavelet details $D_k (k=\mathrm{1,2},3,...,K)$ as shown in Equation (1).

$$\mathsf{\Delta }g(x,y)=A_K+{\sum }_{k=1}^K{D}_k$$

(1)

where the wavelet approximations $A_K$ are the low-frequency background gravity anomalies and the wavelet details $D_k(k=\mathrm{1,2},3,...,K)$ are the high-frequency regional gravity anomalies. The former are generated by the geological bodies in the deep Earth, while the latter are generated by the shallower geological bodies.

The main advantage of wavelet multiscale decomposition in decomposing the Earth’s gravity field is that it has a low-order invariance criterion [30]. As the order of the decomposition increases, the results of the individual lower-order wavelet details do not change. The decomposition result is simply a new pair of wavelet details and wavelet approximations added to the original. Therefore, when we use wavelet multiscale separation to study the Earth’s gravity field, we can simply increase the order of the decomposition calculation according to our needs. Any existing wavelet details $D_k$ can be further decomposed to obtain the advanced order wavelet approximation $D_k{A}_p$ and wavelet details $D_k{D}_t(t=\mathrm{1,2},3,...,p)$, as shown in Equation (2).

$$D_k=D_k{A}_p+{\sum }_{t=1}^p{D}_k{D}_t$$

(2)

where $D_k{D}_t$ can be further decomposed until its corresponding field source depth matches our research objective. The source depths of gravity anomalies could be calculated by radial power spectrum analysis of the bit-field spectrum theory [30].

2.3. Neotectonic Map for Tibetan Plateau and Adjacent Areas

To facilitate the understanding of the relationships between the tectonic setting and lithospheric discontinuities, we present an active tectonic map of the study area based on the previous studies in Figure 1 [36,37,38].

3. Results

3.1. Bouguer Gravity Anomalies at Different Depths

Performing several wavelet decomposition and field source depth estimations for the Bouguer gravity anomaly of the study area, we conclude that the 7th-order wavelet details correspond to a depth of 136 km. This depth is the maximum depth of the area between the lithosphere and the upper mantle of the Tibetan Plateau. Thus, to reveal the characteristics of the gravity field and infer the corresponding tectonic features at different depths in the study area, we performed a 7th-order wavelet decomposition on the Bouguer gravity anomalies using a biorthogonal 3.1 wavelet (bior3.1) and then evaluated the field source depth using the radial power spectrum algorithm. Further, we established the spherical multiquadric (MQ) model [39] on each wavelet detail and calculated the total horizontal derivative (THD), which has a higher accuracy than the vertical derivative (VD) method when dealing with the potential field on grid points [40]. The wavelet decompositions of Bouguer gravity anomalies are shown in Figure 2, and the corresponding field source depths are listed in

Table 1. The estimates of approximate field source depths.

Number of Orders	1	2	3	4	5	6	7
Field depth (km)	0.5	2.5	6.5	10	17	85	136

.

The 1st- to 5th-order wavelet details clearly reflect the tectonic boundaries of the orogenic zone in Figure 2a–e. The Bouguer gravity anomalies of the Pamir Plateau, Tian Shan range and tectonic boundaries of the Tibetan Plateau are very complex. Positive and negative anomalies are alternately distributed and reflect evident features of the tectonic boundaries that separate the Himalayan-Tibetan orogenic zone from its surrounding structures. The complexity is related to the uneven density of the upper crust. As the depth of the field source increases, under intense tectonic activity, widespread metamorphism occurs in the sedimentary structures. Therefore, the gradient belts of Bouguer gravity anomalies connect into larger pieces, and its complexity is reduced. At a depth of 17 km in Figure 2e, the anomalies show a regular distribution between positive values and negative values, which is obviously related to the spatial distributions of major active faults exposed at the Earth’s surface. The elongated belts of high positive gravity anomalies are distributed along the Main Frontal thrust (MFT), Altyn Tagh fault (ATF) and Karakash fault (KKF) on the southern and northern margins of Tibet.

The spatial distribution characteristics of Bouguer gravity anomalies change significantly when the depth increases. The negative anomalies in the bottom of the lithosphere are mostly distributed at the eastern and western ends of the Tibetan Plateau in Figure 2f, such as the MFT, KKF, ATF, southern segment of the Yadong-Gulu fault (YGF), Jiali fault (JlF), the Nujiang fault (NjF) and Burmese fold belt (BFB). Of these, the negative anomalies along YGF, southern BFB and NjF extend from the shallow crust to the deep crust. The volcanic rocks in this depth could not form such spatial patterns. It indicates that the negative anomalies mostly result from crustal stacking or folding of the Mohorovicic discontinuity. Furthermore, in the mid-Himalayas between 75° E and 90° E, the anomaly values are close to zero, while the Bangong-Nujiang suture zone (BNS) and its northern lithosphere show significant positive anomalies. This suggests the existence of the subducted Indian plate.

Compared with the overlying crust, the materials are obviously redistributed and readjusted at a depth of 136 km in Figure 2g. The Bouguer gravity anomalies are significantly related to the spatial distributions of major tectonics. There are two significant positive gravity anomalies in this pattern. One corresponds to the Tarim basin, and the other corresponds to the Indian plate close to the WHS that forms an evident wedging structure. The northern boundary of the latter is the Shyok suture zone (SSZ), and its western boundary is highly in accordance with the Chaman fault (CF) and Quetta valley. The CF is widely recognized as a tectonic boundary between the Indian plate and Eurasian plate in this region. These two positive regions seem to collide with each other, and the NW–SE gradient belt of negative gravity anomalies between them is intensely squeezed. This negative anomaly continuously covers the entire Pamir, western Tian Shan and western Tibet. In the Tian Shan range, this negative gradient belt is cut off by the Talas-Fergana fault (TF), and in Tibet, it tends in the direction of the Karakoram fault (KF), which separates the Pamir Plateau and Tibetan Plateau structurally. Farther east, this negative gradient belt is divided into two narrower belts, the Yalu-Zangbo suture zone (YZYS) and western Jinshajiang suture zone (JS). In addition, the Shan Plateau and Sichuan-Yunnan tectonic zone show significant negative Bouguer gravity anomalies. Near the EHS, the positive anomalies form a narrow belt distributed along the JlF, NjF and BFB, and the southern end of the YGF also exhibits significant positive anomalies.

Figure 2h shows the wavelet approximation of the entire study area, which relates to the material distribution in the deep Earth. Compared with those in the overlying lithosphere, the significant negative Bouguer gravity anomalies further tend to the northeast, and the negative gravity anomalies of the NW Pamir and Himalayan tectonic zones are close to zero. This reveals that crustal and mantle materials are squeezed into the Tarim basin and northern Sichuan-Yunnan tectonic zone, where the characteristics of negative anomalies are evident.

3.2. Edge Recognition on Wavelet Transform Details

Discontinuities in lithospheric materials and active fault distributions at different depths can be clearly recognized by the maximum values of the THD. The THD values (THDs) of the boundary zone between the Indian plate and Eurasian plate are higher than those of the surrounding crust in Figure 3. In the Pamir Plateau, the maximum values are mainly distributed along its boundary structures, such as the Kashgarr-Yeshgar transfer system (KYS), KF, SSZ and western segment of the Pamir thrust system (PTS). The global maximum values are mostly located in front of Nanga Parbat of the WHS. In addition, the Pamir Plateau can be divided into two subunits by the N–S-trending Sarez-Karakul fault (SKF). In western Pamir, the maximum values are more obvious than those in the east, which indicates that the crust in western Pamir is much more broken than the eastern crust. On the Tibetan Plateau, the maximum THD values are mainly distributed within the Himalayan tectonic zone south of the KF and YZYS. The THD values over the Gurla-Mandhata detachment (GM) and along a series of N–S-trending normal faults, such as the Sangri-Nacuo fault (SNF) and YGF, are much higher than those of the crust of the immediate neighborhood. On the eastern boundary of the Tibetan Plateau, the intersection position of the Apalong fault (AplF) and JlF in the EHS as well as the NjF and Xianshihe (XshF) also exhibit evident material discontinuities, which indicate the presence of structural discontinuities along the abovementioned structures.

The Qilian-Nanshan fault (QNT) on the northeastern margin of the Tibetan Plateau appears in Figure 4. Tectonic discontinuities in western Pamir become more significant and widespread. In the mid-Himalayas, the material discontinuities are still concentrated on the GM and N–S normal faults. On the eastern boundary of the Tibetan Plateau, crustal discontinuities along the JlF between the YZYS and Apalong fault (AplF) are significant. The NjF, AnhF, XshF and Longmenshan thrust (LmsT) become more evident than in the shallower crust.

The crustal discontinuities of the Tian Shan range are still not evident except in Aksu, China in Figure 5. The area with maximum THDs in western Pamir is significant, and the gradient belt extends farther southward into the Hindukush area. The discontinuities near the WHS are mostly limited to the south of the Rushan Pshart suture zone (RE). The discontinuities between the Pamir Plateau and the Tarim basin become indistinct. Additionally, crustal discontinuities in the Himalayan tectonic zone are more evident than those in northern Tibet. Only the KKF, ATF and QNF exhibit certain extents of structural discontinuities. In addition, the Anyimaqen-Kunlun-Muztagh suture zone (AMS) between the Qaidam basin and the Tibetan Plateau is partly evident. The values of THD along the N–S-trending normal faults in the mid-Himalayas decrease, except for those in the crust west of the GM and east of the SNF. In eastern Tibet, the crustal discontinuities of the EHS are as evident as those of the WHS. The NjF is still obvious.

The area with maximum THD values in the Pamir Plateau disappears in Figure 6. The current depth is close to the crystalline basement of the plateau. This indicates that the crustal materials of the basement are relatively intact. The areas with maximum THD values are mostly distributed around the WHS and SSZ. The distribution of the maximum THD values in the Himalayan tectonic zone is relatively homogeneous. The differences in crustal discontinuities among the western, middle and eastern Himalayas are not as obvious as those in the overlying crust. The GM and the southern segments of the N–S-trending normal faults exhibit significant crustal discontinuities. In addition, the southern end of the YGF and its eastern structures, such as the YZYS, SNF and MCT, exhibit evident crustal discontinuities. The THD values along the JlF and AplF near the EHS decrease, and only the frontal great bend of the YZYS exhibits a degree of crustal discontinuity. In eastern Tibet, the areas with maximum THDs are distributed discretely at almost the same latitude as the NjF. Further, crustal discontinuities in the western crust of the LmsT are obvious.

In Figure 7, the crustal discontinuities are mostly concentrated in several separate parts of the boundary zone between the Indian and Eurasian plates. On the Pamir Plateau, the areas with maximum THDs in the E–W direction are mainly distributed in front of the WHS, from the SSZ to the KKS. In the Himalayas, crustal discontinuities are more evident near the KF and southern segments of the ALAP, and they are cut by the MCT and eastern SNF. The distribution of areas with maximum THDs obviously decreases in eastern Tibet. Only very limited discrete maximum values are found near the EHS and the NW segment of the NjF.

The spatial distribution characteristics of the THD values at depths of 85 km (Figure 8) and 136 km (Figure 9) are significantly different from those in the overlying crust. This difference can also be identified in Figure 2. Previous research confirmed that the average lithospheric thickness of the study area is 85 km [13]. Thus, the tectonic depths corresponding to Figure 8 and Figure 9 should be the bottom of the lithosphere and the top of the upper mantle. In Figure 8, the areas with maximum THDs are mainly distributed along the eastern and western margins of Tibet. The discontinuity along the Herat fault (HF) and RE in southern Pamir indicates the leading edge of the Indian plate. The discontinuities along the KKF, KF, ATF and XSDF are still evident at current depths, which indicates that these faults are major fracture structures that cut deep into the lithosphere. The lithospheres between the Himalayan tectonic zone and the EHS are obviously separated by the SNF. The areas with maximum THDs of eastern Tibet exhibit significant material discontinuities in the EHS. Combined with Figure 2g, a continuous high positive gradient belt protrudes into the negative anomalies and separates the negative anomalies into several discrete segments.

The source depth in Figure 9 corresponds to the upper mantle in the study area. The spatial distributions of the maximum THD at this depth show significant differences between the Pamir Plateau and Tibetan Plateau. The tectonic discontinuities in the Pamir Plateau are much more significant than those in the Tibetan Plateau. In detail, the lithospheric discontinuities of the Pamir Plateau are mainly distributed north of the MFT and extend northeastward to the southern Tian Shan thrust (STT), KYS and KKF, which form the western boundary of the Tarim basin and indicate the existence of structural discontinuities between the two major tectonic units. Such a pattern of maximum values exhibits the leading-edge characteristics of the subducting Indian plate. In the Tian Shan range, the maximum THDs are mostly distributed west of the TF, which is also revealed by the spatial distribution of the Bouguer gravity anomalies shown in Figure 2g. In addition, the Tajik-Afghan basin and the Tarim basin are relatively intact compared with the Pamir Plateau. On the Tibetan Plateau, the pattern of maximum values is approximately elliptical; the major axes are consistent with the directions of regional tectonic structures and display evident clockwise rotation near the EHS. The most significant characteristic of eastern Tibet is that the maximum values form a gradient belt that is distributed along the Gaoligong fault (GlgF) and Sagaing fault (SF).

3.3. Pamir Plateau and Adjacent Areas

The material discontinuities along the western segment of PTS can be obviously tracked between the SKF and the Tajik-Afghan basin in Figure 3, Figure 5 and Figure 6. Compared with its deeper crust, the crust shallower than 17 km mainly reflects the collision between the Pamir Plateau and Tian Shan range. According to Figure 6 and Figure 7, the basement of the Pamir Plateau is relatively intact since the areas with maximum THDs are primarily distributed along boundary structures such as the PTS, SSZ and WHS, among which the SSZ should be a deep active fault that cuts through the entire lithosphere.

The current activity of the faults can be inferred from the spatial distribution of the area with maximum THDs. There are obvious maximum values distributed along the KF when the depth exceeds 0.5 km in Figure 3, which is not obvious at the surface of the Earth. Thus, it is suggested that the KF is a major fault between the Pamir Plateau and the Tibetan Plateau, but its present-day activity might not be obvious enough because its shallow crust is much less affected by its activity. This indication was proven by GPS observations that the slip rate along the KF is less than 5 mm/yr [41]. The KYS between the Pamir Plateau and the Tarim basin should be a major deep fault that cuts through the entire lithosphere because the material discontinuities along this fault can be recognized in all wavelet details. The RE and KKF are more evident near the bottom of the regional basement but not evident in the shallower crust (as shown in Figure 3 and Figure 4), which indicates that these two faults are inactive.

The SKF separates eastern Pamir from western Pamir. Based on the observed fault scarps, offset geomorphologic features and unconsolidated fault gouge, Strecker et al. (1995) suggested that the SKF accommodates sinistral shear and an approximately east–west extension between these two regions [42]. The eastern Pamir Plateau appears to be largely aseismic, whereas western Pamir shows several seismically active zones. The seismicity pattern of both eastern and western Pamir is significantly consistent with the spatial distribution of THDs in Figure 10. The field depths are also consistent with the deposit thickness within the Pamir Plateau, which ranges from ~3 km to ~6.5 km, according to the model CRUST1.0. Thus, the areas with maximum THDs in the uppermost crust of western Pamir mostly represent the fragmentation of the sedimentary layer. Based on a 2-year seismic record from a local network, Schurr et al. (2014) explained this seismicity pattern by the gravitational collapse of sedimentary rocks from western Pamir to the Tajik-Afghan basin [37]. In addition, the lack of significant seismicity and relatively lower level of THDs in eastern Pamir’s interior indicate that the crust is relatively intact compared with western Pamir. Based on GPS observations, Zubovich et al. (2010) indicated that eastern Pamir is moving northward with tiny interior deformations [43].

The Pamir-Hindukush region is unique in featuring intense intermediate depth seismicity in an intracontinental tectonic setting. From Figure 11b, we could detect two distinct subsets of seismicity beneath the Pamir-Hindukush seismic zone, which are separated by a seismic gap and a 90° change in dip and strike directions close to 36.8° N, 71.4° E. This peculiar geometry of these two species, apparently dipping in opposite directions, invariably leads to the question of their provenance. The seismic gap is located on the boundary between the positive and negative Bouguer gravity anomalies, which is marked by a red inverted triangle in Figure 11a. The opposite signs of Bouguer gravity anomalies along the trace of the profile in Figure 11a indicate their different tectonic provenances beneath Hindukush and southern Pamir. Using multiscale seismic arrival time data acquired from various sources, Bhatti et al. (2018) generated a high-resolution 3D tomographic P-wave velocity model of the crust and upper mantle beneath the NW Himalayas and Pamir-Hindukush [2]. They suggest that the intense intermediate depth seismicity beneath Hindukush is related to the subduction of the lower crust of Indian continental origin, which becomes deeper, steeper and almost vertical from western to central Hindukush. Further north, the Indian plate overturns under southern Pamir upon a close interaction with the Asian plate. Thus, the Asian plate subducts southward beneath southern Pamir, and the southern Pamir slab is mostly made up of Eurasian material [2,44,45].

3.4. Yadong-Gulu Fault

Under the continuous convergence between the Indian and Eurasian continents during the Cenozoic era, there formed a series of normal faults and rifts with approximately north–south trending in southern Tibet. This widespread distributed of the north–south trending rifts is commonly attributed to the east–west extension of southern Tibet, among which the YGF accommodated most of the east–west crustal extension across southern Tibet from 80–90° E [16]. This might be the key structure to understand the mechanism of east–west extension.

The positive Bouguer gravity anomalies are distributed west of the YGF at depths of 10 km and 17 km in Figure 12. As the depth increases, the area of positive Bouguer gravity anomalies increases. The distribution of positive anomalies is consistent with the Nyainquentanglha granite body, which is composed of the young batholith of Miocene intrusive rock [45]. The geophysical observations provided by the INDEPTH project [46] suggest that the Miocene intrusive rock is correlated with the existence of a local melting layer that deepens from ~15 km to ~20 km. A 3D electrical structure across the YGF revealed by magnetotelluric data [45] indicates that the electrical structure of this local melting layer is resistive bodies, which are further intruded by underlying conductors at depth. They also suggest that the tearing of the Indian lithospheric slab at a depth of about 80 km may play a key role in the formation of those high-value conductors. As shown in Figure 12c, the YGF is the boundary between positive Bouguer gravity anomalies in the west and negative anomalies in the east. It indicates that the YGF might be a structural boundary at the depth corresponding to the bottom of the crust and the upper mantle lithosphere. According to the pattern of THDs in Figure 12d, local areas with maximum THDs along with the trace of the YGF suggest that there might be a significant material discontinuous interface beneath the YGF. Tian et al. (2015), using virtual deep seismic sounding, identified a regional Moho uplift located west of the YGF [47]. Liang et al. (2016), using finite frequency tomography, revealed a significant low-velocity zone on the west flank of the YGF in the uppermost mantle, which was interpreted as the fragmentation of the Indian lithospheric slab [48]. Both of these studies suggested discrete asthenosphere upwelling from the fragmentation of the Indian lithospheric slab west of the YGF and thus that it contributes significantly to the lateral discontinuity across the YGF.

On the eastern flank of the YGF, Gang et al. (2017) detected a high-value conductor [49] which is consistent with the distribution of negative Bouguer gravity anomalies shown in Figure 12c. The sub-Moho earthquakes indicate that a high-value conductor relates with the existence of an eclogite layer under the lower crust [48,49]. Guo et al. (2014) detected a thermal origin from the increasing conductivity of the eclogite layer and a laboratory study suggested that the electrical conductivity of eclogite will increase with increasing temperature [50]. Therefore, the negative Bouguer gravity anomalies east of the YGF may indicate the existence of the eclogite layer, and the high conductivity of the eclogite layer may be caused by the effect of upward heat transfer, which could also be indicated by the regional abundant geothermal resources, i.e., Yangbajing geothermal field.

3.5. Eastern Tibet

In eastern Tibet, most faults are concentrated within the shallow crust except for the NjF. The discontinuities along the XshF are only revealed at depths below 6.5 km, which indicates that this fault is active [51] at present. Most faults in northern Tibet cut deep into the crust, and their influences on the material discontinuities in the upper crust are very limited. The areas with maximum THDs distributed in the Himalayan tectonic zone are mostly consistent with a series of N–S-trending normal faults, and the corresponding discontinuities along these faults are widespread in the crust. This pattern indicates that the E–W extension of southern Tibet is a commonly occurring phenomenon. In addition, the crustal discontinuities over the GM and along the SNF are significant and indicate that these two structures cut through the entire lithosphere.

In Figure 8, the LmsT, especially its southern segment, exhibits significant material discontinuities. In Figure 2, the shape of the Sichuan basin can be clearly detected at depths from ~0.5 km to ~85 km, which indicates the different tectonic features between the Sichuan basin and the Tibet Plateau. Especially in Figure 2f, negative Bouguer gravity anomalies are distributed on the western flank of the LmsT and positive Bouguer gravity anomalies are mostly distributed in the Sichuan basin. Such a pattern of Bouguer gravity anomalies is consistent with the deep structural characteristics of eastern Tibet detected by Zhang (2010), using seismic tomography image technology [52]. As shown in Figure 13, there are two profiles, one perpendicular to the trace of the LmsT and another parallel to it. Both of them indicate that the low-speed body is widely accumulated on the western flank of the LmsT, whose depth ranges from ~10 km to ~70 km. Such a low-speed body might be formed by the blocking effect on the material flow in the lower crust by a high-speed body beneath the Sichuan basin. Additionally, the volume of the accumulated low-speed body beneath the southwestern segment of the parallel profile in the bottom plot is larger than that beneath the northeast segment, which is also consistent with the Bouguer gravity anomalies mentioned above.

4. Discussion

When depths exceed 85 km, the tectonic features are very different from those in the overlying crust. In Figure 8, the areas with maximum THDs are mainly distributed around the WHS and EHS. In the Pamir Plateau, the discontinuities of geological materials are mostly distributed between the RE in the north and the CF and MFT in the south. The CF and MFT are two widely recognized boundaries between the Indian plate and the Eurasian plate. So, the areas with maximum THDs in the WHS represent the leading edge of the subducted Indian plate, which has already reached the SW margin of the Tarim basin.

In the upper mantle (Figure 9), areas of maximum THDs are mostly concentrated beneath the Pamir Plateau, the Hindukush, the Sulaiman Range and the WHS and exhibit significant leading-edge tectonic features. Given the contrasts between the EHS and WHS, discontinuities along the WHS are widespread throughout the lithosphere, whereas in the EHS they are mostly concentrated at depths of 85 km and shallower than 10 km. This suggests that the collision between the Indian and Eurasian plates is deeper and more complete in the west and shallower and more discrete in the east. All these findings support the basic assumption that structural closure in the west is the boundary condition for the eastward tectonic escape of the study area. This could also explain the basic mechanisms by which the Pamir Plateau and Tarim basin are now moving northward at the same rate [6,7].

Compression and tectonic closure are important conditions for the formation of eastward escaping deformation in the Himalayan tectonic zone, but the main debate on the mechanism of its tectonic escape lies in the role played by active rupture in this process. GNSS observations [7] show that the present E–W extensional deformation along the Himalayan tectonic belt is relatively unique. The effect of N–S normal faults is negligible. Among these, the YGF and SNF are the key structures that mark the change in horizontal deformation of the Earth’s crust from north–south extrusion to clockwise rotation. The SNF and YGF are two major active faults that cut through the entire lithosphere of all the N–S normal faults in the study area. This suggests that only the active faults that cut deep into the lithosphere can influence the progress of the eastern escape of the Himalayan tectonic zone. Most of the major active faults have little influence on the progress of the eastern escape of the Himalayan tectonic zone.

5. Conclusions

Overall, this study provides detailed patterns of the spatial distribution characteristics of major active faults and structural discontinuities in the crustal lithosphere and upper mantle, revealing the tectonic evolution mechanisms of the Tibetan Plateau and Pamir Plateau.

We found 15 faults that cut through the lithosphere, mostly distributed in the Pamir Plateau, the Himalayan orogenic zone and northern Tibet. They are the Chaman fault, Shyok suture zone, Altyn-Tagh fault, Karakash fault, Karakoram fault, Talas-Fergana fault, Kashgarr-Yeshgar transfer system, Rushan-Pshart suture zone, Sangri-Nacuo fault, Main Frontal thrust, Burmese fold belt, Yadong-Gulu fault, Gaoligong fault, Sagaing fault and Nujiang fault.

In the Pamir Plateau, we have elucidated two tectonic mechanisms of two famous geodynamic phenomena. The first explains why intense seismicity is confined to the western upper crust of the Pamir Plateau. It could be explained as a significant fragmentation of crustal material, mostly caused by the gravitational collapse of sedimentary rocks from western Pamir into the Tajik-Afghan basin. The second is the tectonic signature of intense intermediate-depth seismicity beneath Pamir-Hindukush. Two distinct structures with opposite Bouguer gravity anomalies indicate that they belong to the Indian and Eurasian plates, respectively. In the Tibetan Plateau, the lateral discontinuity of the deep medium is consistent with the mechanism of abundant geothermal resources beneath the Yadong-Gulu fault. The Indian plate is ruptured at the base of the lithosphere, leading to upwelling of the soft flow ring along the rift and intrusion into the lithospheric interior.

Most importantly, we have revealed different tectonic features of the western and eastern Himalayan syntaxes. Tectonic discontinuities in the WHS are widespread at all depths. However, tectonic discontinuities in the EHS are more complex and not well defined between depths of 10 km and 85 km. These tectonic features suggest that the closed structure was only formed to the west. This is the basic mechanism responsible for the tectonic escape of the Tibetan Plateau. Based on this finding, we could conclude that the spatial distribution of crustal horizontal deformation patterns in the Himalayas is related to the interaction between the Indian and Eurasian plates in the deep crust and upper mantle.