3.1 Model residual distributions, errors, and fault properties
Actual GPS measurements inside the model and the simulation results of the corresponding positions showed the same motion trend and a very small deviation in general (Fig. 3a). The simulated and measured residual distributions showed that the residual values were small and disordered (Fig. 3b), indicating that the residual distributions did not contain comprehensive motion information. A total of 220 east and north velocity component residuals were counted, in which the residuals less than 1 mm/yr, 2 mm/yr, and 2–3 mm/yr corresponded to 36.3%, 62.7%, and 80.5% of the total, respectively (Fig. 3c). Errors of less than 5° (80.1%), 10° (82%), and 15° (98.4%) were measured based on the statistics of the directional angle errors between the simulation and actual measurements (Fig. 3d).
The simulation results of fault properties and slip rate values were consistent with the geological results (Table 2), indicating that the simulation results generally reflect the movement and force of regions and faults. It should be noted that the simulation rate of some faults was lower than the geological results, which may be related to the fact that the model is a continuous-medium model and the faults were simplified to easily deformable strips, which inevitably weakened the role of faults in the tectonic evolution of the Tibetan Plateau (Song, 2010). Moreover, the width of some faults in the model was narrower than the actual fault width, and the absorbed deformation was relatively small. Finally, this discrepancy between simulation rate and geological results could be explained by the local sections of faults being the focus of geological surveys, which may make the motions of different fault sections quite different.
Table 2
Statistical table of geological survey results and simulation results of fault activity
| Fault zone | Geological resultz | simulation results | references |
Fault Properties | Slip Rate (mm/yr) | Fault Properties | Strike-slip rate (mm/yr) | Tension and compression rate (mm/yr) |
1 | Karakoram | RL | 14.5 | RL + R | 5.5-9 | 3.8–4.5 | Li et al., 2008; Zheng, 1993 |
2 | PalongCo | R | — | R为主 | 0.2-2 | 5–8 | Liu et al., 2013; Wu et al., 2015 |
3 | GaizedongCo | RL + R | 7.4 | RL + R | 0.5-4 | 0.2-4 | Xiao et al., 2010 |
4 | GyaringCo | RL | 0.92–10.05 | RL + N | 4.9–7.1 | 0.3–2.1 | Yang et al., 2011; Wu et al., 1994; |
5 | BengCo | RL + N | — | RL + N | 1.8-6 | 0-2.9 | Wu et al., 1994; Yang et al., 2011; Ning, 2006 |
6 | Jiali | RL + R | Northwest segment:6–8 Southeast segment:3 | RL + R | West segment:3–6 Middle: 0.5–1.5 East segment: 3-4.3 | 0-1.2 0–2 0–3 | Ren et al., 2000; Song et al., 2013; |
7 | PalongCo | LL + N | — | N + LL | 0.5–4.2 | 1.5-5.0 | Elliot et al., 2010; Styron et al., 2010; Qiu et al., 2010 |
8 | DangyeYongCo | N | — | N + LL/RL | 0-2.5 | 1.5–5.6 | — |
9 | Jiagang-Dingjie | N + RL | 4 | N + RL | 0.3-1 | 3.9–6.1 | Gan W et al, 2007 |
10 | Nyainqentanggula –Yadong-Gulu | NS segment :RL + N NE segment :LL + N | North segment:3.5–5.5 Middle:7.5–9.5 | NS segment RL + N, NE segment LL + N | North segment:2.9–5.4 Middle:1.1–5.1 South segment:4.9-7.0 | North segment:2.2–5.7 Middle:5.5–9.1 South segment:2.1–4.5 | Wu et al., 1990; Chen et al, 2004 |
11 | Magnion | — | — | R + LL | 0.7-5 | 5-7.7 | — |
12 | Himalayan | R | 14–22 | Mid-west segment:R + RL East segment:RL | 3-9.9 | 2-10.1 | Song, 2010; Li, 2019; Zhang, 2013; Wang., 2011 |
Note: — Indicates no result; LL: left- lateral strike-slip, RL: right- lateral strike-slip, R: thrust fault, N: normal fault |
3.2 Regional deformation field, stress field distribution, and fault influence
The velocity field vector and contour distribution provided by the simulation in the research area showed a very fast regional velocity field of 13–36 mm/yr with a highly uneven spatial distribution (Fig. 3a). The velocity gradually decreased from south to north and diverged from the center to the east and west. Moreover, the velocity field of the Himalayan tectonic belt was ~ 36 mm/yr, and the northern border of research area exhibited a motion of 13–25 mm/yr, which was shortened by 11–23 mm/yr in the north to south direction. Bounded by 90° E (near the Nyainqentanggula-Yadong-Gulu fault), the eastern region velocity was ~ 36 mm/yr in the middle and ~ 18.6 mm/yr in the eastern limit, with a maximum decrease close to 17 mm/yr. The direction rotated clockwise from NE to near EW, with the western region gradually decreasing to 36 mm/yr in the middle and ~ 15.1 mm/yr in the northwest, showing a maximum decrease of 21 mm/yr and shifting the direction from NE to near NS. These results reflect the NS-trending of compression deformation and the EW-trending of extensional deformation in the research area. The EW-trending extension was particularly strong in the area east of the Nyainqentanggula-Yadong-Gulu fault, and the plateau material was extruded rapidly eastward, which is consistent with the numerical simulation results of Wang et al. (2006).
Figure 4
The calculated stress field obtained in this study resulted from the stress rate under a steady state. The regional principal stress simulation showed a principal compressive stress direction changing from west to east to near NS to NE (Fig. 4). The magnitude of the principal extensional stress in the area was generally lower than the principal compressive stress, indicating that the compressional deformation was predominant and accompanied by tensile deformation. The ratio of principal extensional stress/principal compressive stress varied from region to region. The extensional stress in the southern region was ~ 1/3 of the principal compressive stress, and the ratio near the Qiangtang block was ~ 1/1. These results are consistent with the principal strain field results obtained using GPS (Zhang et al., 2003; Wu et al., 2020; Qu et al., 2021). In summary, the stress field in this region was inhomogeneous.
The faults effect can be clearly seen in the velocity contour distribution (Fig. 3a). The velocity contours were bent on both sides of the EW-trending and NW-trending faults, indicating that the NEE movement on the north side of the fault was slower than that on the south side. On both sides of the internal NS-trending normal faults, particularly the DangreYongCo and Nyainqentanggula faults, the velocity contour curve showed that the fault's east side movement was slower than that on the west side. This is observable because the fault has the ability to absorb deformation, which has a particular hindering effect on the northward movement and the EW direction extension of the research area, changing the movement pattern to a certain extent. Additionally, the stress on the fault's south side at the southern boundary of the Lhasa block was significantly higher than that on the north side (Fig. 4), indicating that the fault hinders the stress transfer and promotes local high-stress accumulation.
3.3 Correlation of horizontal slip rate and stress distribution of faults with earthquakes
3.3.1 Movement and force characteristics of the NW-EW trending fault zones on the southern and northern boundary of the Lhasa block
Figure 5
Overall, the simulation results of the horizontal tensile-compressive rates of vertical faults, strike-slip rates of parallel faults, and corresponding normal stress and shear stress indicated that the movement and forces of the three main fault zones on the southern boundary were different and uneven (Fig. 5). The Karakorum fault zone was subjected to significant right-lateral shear and compression, dominated by right-lateral strike-slip with thrust movement. The strike-slip rate was 5.5–9.0 mm/yr, and the compressive rate was 4.3 mm/yr. As the western border of the Qinghai-Tibet Plateau, the Karakorum fault played an essential role in the eastward plateau movement. Furthermore, the middle segment of the Himalayan fault zone absorbed a large amount of convergent deformation, and the compression rate gradually decreased from 10.1 mm/yr in the middle to ~ 2 mm/yr in the west and east ends, and the strike-slip rate decreased from 9.9 mm/yr in the middle to 5 mm/yr and 3 mm/yr in the west and east ends, respectively. The most significant compression rates were observed in the central and western sections. Contrastingly, the largest strike-slip rate was found in the eastern segment. These findings indicate that the northern Himalayan orogenic belt is now dominated by extensional structures, which is consistent with geological findings (Zhang, 2013). In terms of stress, the middle and eastern segments were subjected to a relatively strong compression and strong right-lateral shear, respectively, reflecting the Himalayan orogenic belt's present deformation coexistence with the NS-trending shortening and EW-trending tension, and the tension activity was strongest in the eastern segment of the fault. The focal mechanism solution (Xu et al., 2005) showed that the central segment of Himalayan fault is dominated by thrust-type earthquakes, whereas strike-slip earthquakes dominate the eastern segment. Nonetheless, relevant geological research on the Magnion fault zone is relatively scant. The simulated strike-slip and compressive rates were 0.7–5.0 mm/yr and 5.0–7.7 mm/yr, respectively, indicating a significant left-lateral strike-slip and tensional normal fault characteristics, which is consistent with the numerical simulation obtained by Bao (2012) and others.
The fault activity of the northern boundary of the Lhasa block is closely related to its strike. For instance, the NWW- and EW-trending faults showed right-lateral strike-slip and compression movements, whereas the NW-trending faults showed right-lateral strike-slip and tension movements. Meanwhile, the faults exhibit significant differences in segmentation. For example, in the BangongCo and Gerze-DongCo fault zones, an interphase distribution dominated by compression and strike-slip was observed along the fault, and the fault segment with significant compression had a weak shear. These results reveal that the northern fault zone of the Lhasa block is an important adjustment zone for NS-trending compression and EW-trending extension. However, the force and rate values differed significantly in different fault sections. For instance, simulated strike-slip rates of the northwest, middle, and southeast segments of the Jiali fault were 3.0–6.0, 0.5–1.5, and 3.0–4.0 mm/yr, respectively, i.e., the strike-slip rate decreased and then gradually increased from west to east along the fault. Previous research results have also shown the above movement features (Ren et al., 2000; Shen et al., 2001; Zhang et al., 2003). The slip rate of the Jiali fault was mostly lower than the 10–20 mm/yr suggested by the "extrusion-escape model" (Armijo et al., 1989), and higher than the 2–3 mm/yr suggested by the "horizontal shortening thickening model" (England et al., 2010). The uplift model of the Qinghai-Tibet Plateau was very complicated, which may be between the "extrusion-escape model" and the "horizontal shortening thickening model,” which is consistent with the views of Li et al. (2021).
The NW-trending GyaringCo and BengCo faults are large-scale right-lateral strike-slip faults and strong seismic activity belts in southern Qinghai-Tibet (Wu et al., 1994; Yang et al., 2011). These two faults are located at the intersection of the NS rifts and the EW fault, and the pull-apart basin is the main tectonic landform in this region (Wu et al., 1994; Ning, 2006). The simulation revealed that the GyaringCo and BengCo faults were dominated by right-lateral strike-slip, in which their eastern segment exhibited a 6–7 mm/yr and 3–6 mm/yr strike-slip, respectively. Similar results were obtained in the geological survey performed by Yang et al. (2011).
Finally, the dextral and normal motion characteristics of the fault were closely related to its background stress field and strike characteristic. Compared with other near-EW-trending faults on the northern boundary of Lhasa block, NW-trending faults were more affected by NS-trending compression and NS-trending rifts.
3.3.2 Movement and force characteristics of the NS-trending normal faults and their relationship with earthquakes
The NS-trending fault zones in the central part of Lhasa block were mainly strike-slip motion normal faults, with an extension rate of 1.1–7.2 mm/yr (Fig. 5). The total extension rate of the four normal faults was 11–18 mm/yr, indicating their strong ability to absorb EW-trending deformation. The principal stress distribution of NS-trending normal faults showed that most principal tensile stress was significantly higher than the principal compressive stress; the direction was basically perpendicular to the fault orientation, and the magnitude of the eastern faults was the most prominent (Fig. 6). The motions and forces of the fault were generally consistent with the fact that the focal mechanism solutions for strong earthquakes in this region are primarily normal fault types.
Figure 6
The simulation showed that the movement and force distribution of the Nyainqentanggula-Yadong-Gulu fault zone were related to the fault strike. With respect to the fault properties and activity intensity, its northern segment showed an NE-trend, with tensional and left-lateral strike-slip rates of 2.2–5.7 mm/yr and 2.9–5.4 mm/yr, respectively. Moreover, its middle segment showed a near NS-trending, with tensional and right-lateral slip rates of 5.5–9.1 mm/yr and 1.1–4.1 mm/yr, respectively, with a significant extensional movement; its southern segment showed an NNE-trend, with tensional and right-lateral slip rates of 2.1–4.5 mm/yr and 4.9–7.0 mm/yr, respectively. Overall, the strike-slip rate first decreased and then increased from the fault's north to south, while the extension rate first increased and then decreased, which is consistent with the geological results obtained by Armijo et al. (1986) and the GPS observation of Zhang et al. (2003).
In relation to the tensile stress accumulation rate, the NE-trending and NS-trending sections of the northern segment of the fault exhibited a 200–400 Pa/yr and 600–900 Pa/yr, respectively, the middle segment 800–1600 Pa/yr, and the southern segment 300–500 Pa/yr. These results indicate that tensile stress accumulates faster in the NS-trending sections of the middle and northern segments of the fault. Notably, the direction of principal stress changed with the fault's strike. The maximal principal stress in the middle and southern segments of the fault zone was the principal tensile stress, and the direction was the near EW direction perpendicular to the fault. The maximal principal stress in the NE-trending section of the north segment was the principal compressive stress, with a direction angle of 30–60° relative to the fault, which was conducive to the promotion of left-rotation shear and weak-tensile motion.
In summary, in the middle NS-trending sections of the Nyainqentanggula-Yadong-Gulu fault, the extension rate and accumulation rate of tensile stress were relatively large, the principal tensile stress was dominant, and the tensile force and fault activity intensity were highly significant. There have been seven normal-fault-type strong earthquakes (M ≥ 5) since 1976 in and around middle of the Nyainqentanggula-Yadong-Gulu fault segment, such as the 2008 Dangxiong M6.6 earthquake.
Similar to the Nyainqentanggula-Yadong-Gulu fault zone characteristics, the DangreYongCo fault also showed characteristics of fault movement related to the strike. The NNW segment strike showed a right-lateral slip, while the NS and NNE segments showed left-lateral slips, whereas these two slip types were alternately distributed along the fault. In contrast, the tensile rate in the NNW-trending fault section was 1.1–2.3 mm/yr and the tensile stress accumulation rate was 200–630 Pa/yr, while the tensile rate in the NS and NNE sections was 3.8–4.9 mm/yr and the tensile stress accumulation rate was 620–1200 Pa/yr. In general, the extensional movement and tensile stress of the NS-trending and NNE-trending sections of the northern segment of the fault were the most significant.
Compared with other normal faults, the Jiagang-Dingjie fault zone had a larger tensile rate of 3.9–6.1 mm/yr and was subjected to significant tensile stress, with a cumulative rate of 680–1300 Pa/yr. However, the segmentation characteristics of the faults were not obvious. According to the historical earthquake statistics, moderate and strong earthquakes have occurred frequently throughout the fault and its vicinity, and the spatial distribution of the epicenters has been relatively uniform. The overall tensile rate and tensile stress of the PalongCo fault zone were not high, but the segmental characteristics were significant. Compared with the northern segment, the central and southern segments of the fault were more active, with left-lateral strike-slip and tensile components of 3.0–4.2 mm/yr and 3.3–5.0 mm/yr, respectively. Moderate-strong earthquakes (M ≥ 5) with magnitude of five and above were primarily concentrated in this fault segment.
From the above analysis, it can be seen that the fault movement and stress simulation showed a noticeable difference in the mechanical properties and movement patterns between faults and fault segments. Some fault segments, such as the central-southern segment of PalongCo (NNE-trending), the NS- and NNE-trending fault segments of DangreYongCo, and the middle NS-trending segment of the Nyainqentanggula-Yadong-Gulu and Jiagang-Dingjie faults, often showed a high tensile stress accumulation rate, and the direction of the principal tensile stress was nearly EW-trending and perpendicular to the fault segment near NS-trending, which was conducive to the breeding and occurrence of normal fault-type earthquakes. Therefore, moderate-strong earthquakes (M ≥ 5, since 1976; or M ≥ 6.5, overall) were concentrated on these fault segments. The above results indicate that the fault segments with high tensile rates and near NS direction controlled moderate-strong earthquakes to some extent.