Imaging the crustal structure beneath the eastern Tibetan Plateau and implications for the uplift of the Longmen Shan range

Earth and Planetary Science Letters 379 (2013) 72–80

Contents lists available at ScienceDirect

Earth and Planetary Science Letters www.elsevier.com/locate/epsl

Imaging the crustal structure beneath the eastern Tibetan Plateau and implications for the uplift of the Longmen Shan range Xiaoyu Guo a,b , Rui Gao a,∗ , G. Randy Keller b , Xiao Xu b , Haiyan Wang a , Wenhui Li a a b

Institute of Geology, Chinese Academy of Geological Sciences, Beijing, 100037, China School of Geology and Geophysics, The University of Oklahoma, Norman, OK 73019, USA

a r t i c l e

i n f o

Article history: Received 25 March 2013 Received in revised form 31 July 2013 Accepted 1 August 2013 Available online 28 August 2013 Editor: P. Shearer Keywords: eastern Tibetan Plateau Longmen Shan block deep seismic reflection profile crustal structure uplift of the Longmen Shan range

a b s t r a c t Competing models have been proposed to account for the unusual uplift of the Longmen Shan fault zone in the eastern Tibetan Plateau and resulting hazards. However, due to the terrain and thick sedimentary cover, the crustal structure of the eastern Tibetan Plateau remains uncertain. Therefore, testing these models has been difficult. In 2011, a 310 km long, SE-trending SinoProbe-02 deep seismic reflection profile was recorded in easternmost Tibet in order to study its regional crustal structure. The resulting image revealed the detailed crustal structure of eastern Tibet, which when combined with geological, GPS (Global Positioning System), and geochemical evidence, strongly suggests that Yangtze sub-continent crust extends beneath the region. The seismic profile also images the tremendously thick Triassic sedimentary cover in the Songpan-Ganzi terrane (SGT). These Triassic sediments vary considerably in thickness across several crustal blocks. In addition, both the Longriba fault zone of the northeastern SGT and the Longmen Shan fault zone show strong intracrustal reflections that terminate at a depth coinciding largely with the crust–mantle boundary (Moho). Accordingly, we propose a new tectonic model based on an integrated analysis of this seismic reflection profile and previous GPS measurements. In this model, crustal-scale deformation is suggested to have participated in the oblique extrusion and uplift of the easternmost edge of the Tibetan Plateau along the Longmen Shan. In a broader context, the lithospheric configuration imaged by the seismic reflection profile will advance our understanding of the tectonic response of the eastern Tibetan Plateau to the ongoing India–Eurasia collision and has new implications for the estimating seismic risk in the region. © 2013 Elsevier B.V. All rights reserved.

1. Introduction On 12 May 2008, the devastating Mw 7.9 Wenchuan Earthquake struck the Longmen Shan (LMS) thrust belt along the eastern margin of the Tibetan Plateau (Fig. 1a). This was the largest destructive earthquake in China in recent years. It caused ∼80,000 fatalities and physical damages of ∼$100 billion US. Although some researchers have recognized the long-term regional hazards (Densmore et al., 2007), the underestimated seismic risk and poor preparation for a large earthquake were attributed due to the relative quiescence of seismic events on the LMS fault zone (Figs. 1a and 1b). The LMS fault zone consists of a series of northwest dipping thrust faults, including the Wenchuan-Maowen fault in the northwest, the Beichuan fault in the middle and the Pengguan fault to the southeast. The Pengguan fault merges with the Beichuan fault and the Beichuan fault roots into a deep detachment at 15–17 km depth (Li et al., 2010). How far the WenchuanMaowen fault extends downward, so far, remains uncertain.

*

Corresponding author. E-mail address: [email protected] (R. Gao).

0012-821X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2013.08.005

Even today, the cause of the LMS thrust-Nappe belt uplift and resulting large earthquakes remains uncertain. There are at least six end-member models proposed regarding the unusual uplift of the LMS fault zone (see review by Yin, 2010); (1) channel flow in the mid and lower crust (Royden et al., 1997, 2008; Zhao et al., 2012) (Fig. 2a), (2) upper-crustal deformation associating with a series of detachments (Hubbard and Shaw, 2009; Hubbard et al., 2010) (Fig. 2b), (3) pure-shear deformation of the whole Tibetan lithosphere (Robert et al., 2010; Yin, 2010) (Fig. 2c), (4) simpleshear shortening of the whole Tibetan lithosphere (Yin, 2010) (Fig. 2d), (5) uplift of the eastern Tibet is associated with westward underthrusting of the Yangtze crust beneath (Clark et al., 2005; Jiang and Jin, 2005) (Fig. 2e), (6) crocodile-type accommodated by indentation of rigid Yangtze crust into the weak SGT (Cai et al., 1996; Zhang et al., 2004) (Fig. 2f). Different models were proposed partly due to the uncertain crustal structure of the eastern Tibetan Plateau that is in general covered by the thick Triassic flysch cover (5–15 km) (Nie et al., 1994; Zhou and Graham, 1996). Testing these models has therefore been difficult. Although many primarily geological and geochemical studies have indicated the possible existence of the Yangtze sub-continental block (YB) beneath eastern Tibet (Burchfiel et al., 1995; Zhang et al., 2006;

X. Guo et al. / Earth and Planetary Science Letters 379 (2013) 72–80

73

Fig. 1. (a) Topographic map of the eastern Tibetan Plateau and Sichuan basin, showing the location of SinoProbe-02 deep seismic reflection profile as a purple line. GPS velocity vectors (Shen et al., 2005, 2009; Zhu and Zhang, 2010) are plotted relative to the Yangtze block. (b) Generalized geological map of the research area based on the 1:2.5 million scale geological map of China. (c) Topographic relief along the seismic reflection line; SGT: Songpan-Ganzi Terrane; WMF: Wenchuan-Maowen Fault; BCF: Beichuan Fault; QCF: Qingchuan Fault; PGF: Pengguan Fault; LRBFZ: Longriba Fault Zone; LRQF: Longriqu Fault; MEF: Mao’ergai Fault; MJF: Minjiang Fault; LMSFZ: Longmen Shan Fault Zone; SB: Sichuan Basin.

Fig. 2. Previously proposed tectonic mechanism to account for the unusual uplift of the Longmen Shan fault zone; (a) the lower crustal flow model; (b) the upper-crustal deformation model; LMS: Longmen Shan; (c) Pure-shear model; (d) Simple-shear model; (e) Underthrusting of the Yangtze crust beneath the Songpan-Ganzi terrane; (f) Indentation of the Yangtze crust beneath the Songpan-Ganzi terrane.

Roger et al., 2010), the westward extent of the Yangtze crust and crustal structure of eastern Tibet have never been fully imaged. Thus, an important constraint on the uplift mechanism of the LMS thrust belt and subsequent natural hazards is lacking. The purpose of this study is to provide needed information on crustal structure beneath eastern Tibet, which is essential to advance our understanding on the tectonic responses to the ongoing Himalayan orogeny and to assess natural hazards. 2. Tectonic setting During the early Mesozoic, prior to the Indosinian orogeny (Late Permian – Early Jurassic), the Songpan-Ganzi terrane (SGT) was originally a remnant ocean that was filled with thick Triassic flysch derived from adjacent orogenic belts (e.g., Yin and Nie, 1993). This remnant ocean was juxtaposed against the passive margin of the YB to the east with onlapping deposition of Triassic fly-

sch onto the margin of the YB (Harrowfield and Wilson, 2005). During the onset of the Indosinian orogeny, the Songpan remnant ocean and its sedimentary fill experienced intense deformation because it was trapped between coeval subduction zones in the north and southwest (e.g. Reid et al., 2005, 2007; Roger et al., 2011). Triassic lithospheric delamination occurred (Zhang et al., 2007) after the crust of the NE Tibetan Plateau thickened to 45 ± 5 km (Lease et al., 2012). As a result, Triassic syn-tectonic adakitictype granitoids are widely distributed in the eastern SGT (Fig. 1b), which are likely sourced from the partial melting of an underlying Proterozoic basement that is part of the YB (Roger et al., 2010; Zhang et al., 2006). In addition, as a convergent zone between the SGT and YB, the LMS thrust-Nappe belt was initiated in the Late Triassic as a sinistral transpressive fault zone (Burchfiel et al., 1995) (Figs. 1a and 1b) due to the compression of the SGT and contemporaneous southeast-directed crustal shortening and thrusting towards the western passive margin of the YB (Roger et al., 2004,

74

X. Guo et al. / Earth and Planetary Science Letters 379 (2013) 72–80

2010). The eastern SGT experienced low cooling between the Late Jurassic to early Cenozoic, indicating the lack of major tectonic events after the amalgamation of the SGT and YB (Roger et al., 2011). Global Positioning System (GPS) measurements indicate that Tibetan Plateau is currently undergoing eastward block motion due to the ongoing collision between the Indian plate and Eurasian plate. In addition, GPS measurements detect a dramatic decrease in the eastward extrusion velocity across the Longriba fault zone in eastern Tibet (Shen et al., 2005, 2009; Gan et al., 2007) (Fig. 1a). The Longriba fault zone consists of two branches. One is the Longriqu fault (LRQF) in the west and the Mao’ergai fault in the east (Xu et al., 2008) (Fig. 1). Further eastward at the LMS range front, no significant eastward convergence is observed relative to the western Sichuan basin (Shen et al., 2005, 2009; Gan et al., 2007). Even though, the LMS mountain belt rises >3 km above the Sichuan basin over a distance less than 100 km and marks a pronounced boundary between the Sichuan basin and eastern Tibet (Fig. 1c). In addition, slip rates on the eastern boundary are much lower than those on the northern and southwestern boundaries of the SGT (e.g. Chen et al., 2000; Densmore et al., 2007; Kirby et al., 2010). No seismic events of magnitude 7.0 were historically recorded on the LMS fault zone before 2008 (Nalbant and McCloskey, 2011). These observations underscore the importance of better information on detailed crustal structure across this region as a key to understanding the mechanism behind the unusual uplift of the LMS range and related natural hazards.

Table 1 Seismic reflection data processing flow. Processing modules

Parameter

Parameter values

Filter

Bandpass Post-stack (shallow) Post-stack (medium to deep)

6–8–100–200 Hz 8–12–45–55 Hz 6–8–40–50 Hz

Static corrections

Reference altitude Replacement velocity

3500 m 4000 m/s

Amplitude recovery

Decibels Time window

0.5 18 000 ms

Velocity spectrum

CDP interval CDP number of traces

50 m 11

Automatic residual statics

Maximum static correction allowed

36 ms

Mute

Top mute Stretching distortion

Standard approach Manually picked

Stack technique

DMO

Migration

2D wave equation migration by finite difference with absorbing boundary conditions

45◦ maximum dip

3. Seismic data acquisition and processing In this paper, we provide an image of crustal structure beneath easternmost Tibet based on new controlled-source seismic data and employ these data to provide new insights on the mechanism of the intense LMS uplift. These seismic data were collected as part of China’s SinoProbe Project, a leading Chinese research effort to investigate the deep structure across China (Dong et al., 2011). In the autumn of 2011, the SinoProbe-02 working group, a consortium from the Institute of Geology at Chinese Academy of Geological Sciences and the University of Oklahoma, USA collected a 310 km long deep seismic reflection profile roughly extending NW-SE from the SGT to the Sichuan basin (Figs. 1a and 1b). Seismic data were acquired by using French SERCEL 428XL recording system and 24-bit digital geophones. In order to obtain high-resolution seismic images of the entire lithosphere, we employed three types of explosive sources with charge sizes of 24 kg, 96 kg and 500 kg, respectively. The 24-kg shots were placed in single shot holes whose depths were 25 m and were closely spaced at 250 m intervals. The 96-kg shots were placed in a cluster of two shot holes at a depth of 25 m, and these clusters were placed at 1000 m intervals. Finally, the 500-kg shots were placed in clusters of 5 shot holes at intervals of 50 km. In the recording, 600 receiving traces were employed with geophone group spacing of 50 m. They were deployed on both sides of the shots with 300 on each side. Thus, the minimum and maximum shot-receiver offsets were 25 m and 14.975 km, respectively. The recording time was 30 s for 24 kg and 96 kg sources and 60 s for the 500 kg sources. This recording geometry for the 24 kg and 96 kg shot sources provided 60 fold common mid-point (CMP) coverage for the processing. Data processing was performed with a combination of commercial software but the Omega system was primarily employed. We ultimately improved the image quality by testing numerous procedures and parameters. The main processing steps include multiple static correction methods, surface consistent amplitude recovery, multi-domain pre-stack noise attenuation, dip-moveout (DMO) correction, CDP stack, and post-stack gain balance, as well as timedomain migration applied to the stacked data. These processing

Fig. 3. Deep seismic reflection profile extending from the Songpan-Ganzi terrane across the Longmen Shan block to the Sichuan Basin (see Fig. 1 for location and see supplementary material for the higher resolution image) (no vertical exaggeration and 6 km/s average seismic velocity is assumed). LRQF: Longriqu Fault; MJF: Minjiang Fault; WMF: Wenchuan-Maowen Fault; BCF: Beichuan Fault; PGF: Pengguan Fault; YB: Yangtze Block; SGT: Songpan-Ganzi Terrane.

steps were designed to preserve the relative true amplitudes. The main steps that were followed during seismic data processing are summarized in Table 1. Thus, these high-density and near-vertical reflection data are the basis of our integrated study of the seismic structure of the crust in the area of investigation. 4. Description of the SinoProbe-02 seismic reflection profile 4.1. Overview The near-vertical reflection section (Fig. 3) (please see supplementary material for a high-resolution image) extends to about 20 s of two-way time (t.w.t.), which corresponds to a depth of ∼50 km from the LMS eastward. The Sichuan basin crust is slower on average and is thus thinner (e.g. Zhang et al., 2009). This seismic image reveals the geometry of the thick Mesozoic sedimentary cover as suggested by previous studies (Nie et al., 1994; Zhou and Graham, 1996) (Fig. 3) and a thick crust. It predominantly reveals two reflection boundaries that are clear but discontinuous. The upper reflector is the top of the basement (crystalline upper crust) that is as deep as 8 s under the Ruo’ergai basin, and the lower one is the Moho (∼16 s). The top of the basement varies considerably along the profile (8 s to 2 s). The Moho is characterized by prominent northwest and southeast dipping highamplitude reflectors between 14 and 18 s (t.w.t.). This result is extremely valuable for providing constraints on models of tectonic

X. Guo et al. / Earth and Planetary Science Letters 379 (2013) 72–80

deformation and orogenic processes in the eastern Tibetan Plateau and along the Longmen Shan Range. The seismic reflection profile can be divided naturally into three domains based on geologic correlations with lateral variations and offsets of events identified in the seismic reflection data (Fig. 3). The subdomains we have identified are the Ruo’ergai basin (northwest of the Longriba fault zone), the LMS block (the zone between the Longriba fault zone and the LMS fault zone), and the Sichuan basin (southeast of the Pengguan fault zone). The topographic surface in the Ruo’ergai basin has an unusually low relief, but high relief appears after crossing the Longriba fault zone to the southeast (Figs. 1c and 3). Here, we define the LMS block as a convergent zone between the eastern SGT and the YB (Figs. 1a and 1b). As shown in Fig. 3, the Ruo’ergai basin is characterized by several packages of southeast dipping reflectors that extend down to 8–10 s (t.w.t.) or ∼20 km (based on an average crustal velocity of 5.5 km/s due to the thick sediments). These dipping reflectors terminate approximately at the surface location of the Longriba fault zone. East of this fault, the sedimentary rocks gradually thin eastward across the LMS block. Under the Ruo’ergai basin, the Moho reflection is discontinuous and each segment dips to the southeast. The LMS block contains two reflection boundaries, which exhibit significant lateral variability in reflection continuity and character. The top interface of the basement (crystalline crust) beneath the LMS block is strongly reflective and rises to the south. Just southeast of the Longriba fault zone, the basement is identified by subhorizontal events that lie at ∼8 s (t.w.t.) below the surface. Further to the southeast, the basement appears to be cut by high-angle faults, and the segments between these faults dip to the northwest. The Moho is delineated by a series of short strongly reflective segments with variable length and continuity at 17 s (t.w.t.) that are offset with a uniform northwest dip. Using an average crustal velocity of 6 km/s, this t.w.t. indicates a crustal thickness of 50 km, which is consistent with the nearby receiver function study of Zhang et al. (2009). The seismic reflection data do not show strong crustal thinning under the Sichuan basin that Zhang et al. (2009) have shown, but a low (∼5 km/s) average crustal velocity due to the thick sedimentary fill would indicate a crustal thickness of ∼37 km, which is in agreement with the result of Zhang et al. (2009). Continuously across the LMS block to the Sichuan basin, there is a decrease in reflectivity of the top interface of the basement. In the Sichuan basin area, the laterally horizontal reflectors are consistent suggesting that the crust of this area is relatively rigid. The Moho beneath the Sichuan basin is relatively flat but is disrupted near the LMS fault zone. 4.2. Seismic sections in areas of structural complexity Four zoomed-in seismic sections were constructed to highlight the structural complexity in the domains of interest (Figs. 4A and 4B). They show reflective events of the Ruo’ergai basin, the Longriba fault zone, the middle section of the seismic reflection profile, and the LMS fault zone, respectively. Panel I is focused on the area beneath the Ruo’ergai basin (Fig. 4A). This panel includes the southeastern portion of the record section from the recent Ruo’ergai basin seismic reflection experiment (Wang et al., 2011). We included these published results because they provide a higher-resolution seismic image. This seismic panel reveals a wide range of intracrustal reflectors and variations in crustal reflectivity. Dipping reflectors are high-amplitude, linear events that in several areas extend through the entire crust and are either southeastward dipping or convexdownward in shape. In a few instances, these events exhibit a relatively subhorizontal attitude with shallow dips in the lower crust

75

that gradually steepen in the middle and upper crust. Comparable to other portions of this seismic reflection profile, the entire upper crust beneath the Ruo’ergai basin is so highly reflective that a discrete boundary making the bottom of the basin cannot be clearly recognized in the reflection data. In panel II of Fig. 4A, two groups of variability in reflectivity of the upper crust exist. The southeastern segment contains horizontal reflectors in the mid-crust but the northwestern segment contains on southeast dipping reflections in the middle crust. Between them, a band of northwestward dipping events is observed, which extends from the Moho well up into the upper crust and possibly to the surface. The dip gradually decreases with depth as it is traced from the upper crust to near the Moho. There is evidence that these reflections flatten out near the crust–mantle boundary. These reflectivity events suggest a fault interface coinciding with the Longriba fault zone. Thus, we interpret the Longriba fault zone to be a boundary, separating two distinct crustal domains. Panel III (Fig. 4B) illustrates the middle section of the seismic reflection profile. The LMS block in this image has a distinct structure, containing both subhorizontal reflectors in the mid to upper crust and more steeply dipping reflectors in the lower crust. Two packages of high-amplitude, discrete subhorizontal reflectors at 4 s (t.w.t.) and 8 s (t.w.t.) are associated with two distinct linear or convex-downward-dipping reflectors. Both are related to the reflectivity of the top surface of the basement due to complex shortening of the LMS block. In the eastern portion of this image, weakly reflective events appear to cut through the mid-crust and extend into the lower crust, and, in a few cases, appear to cut through the Moho. We interpret the main feature in this complex to be the Minjiang fault. The Moho is associated with a banded sequence of reflectors that dip toward the northwest with several offsets. Some of the dipping events that offset the Moho appear to fade out in the acoustically transparent mid-crust. Panel IV crosses the LMS fault zone into the Sichuan basin (Fig. 4B). The base of the sediments beneath the western Sichuan basin is evident on the seismic section as a strong horizontal reflector lying at 6 s (t.w.t.). This horizon is associated with a major detachment connecting both the Beichuan fault and the Pengguan fault (Hubbard and Shaw, 2009; Li et al., 2010; Lu et al., 2012) and merges to the west with a group of northwest dipping reflectors. This group of steeply northwest dipping reflections is interpreted as a complex of thrust faults in the upper crust and may extend as ductile shear zones downward to the Moho. We interpret the main feature in this complex to be the Wenchuan-Maowen fault. In addition, bidirectional-verging faults (southeast and northwest) in the upper crust characterize a pop-up structure. Complex Moho offsets are evident at 14 s (t.w.t.) (∼42 km) beneath the Sichuan basin; northwest dipping Moho offsets are observed between 16–18 s (t.w.t.) (∼48–54 km) beneath the LMS block. These offsets are evidence that the LMS block is highly shortened and that the rigid western Sichuan basin (YB) is resistant to the crustal shortening of the LMS block. We interpret this deformation as a manifestation of the SGT and YB interaction in the Mesozoic, in which the early Mesozoic thrusting offset both the Moho and structures in the upper crust of the western margin of the YB. However to some degree, these features have been reactivated since the Late Quaternary. 4.3. Interpreted seismic profile Fig. 5 is an interpretation of the regional crustal structures. Within the LMS block, estimates of Moho depths are based on reflection times and the velocity results of the corresponding seismic refraction profile. The overall seismic reflection image displays a partly bivergent crustal architecture in the LMS block that lies between the Sichuan and Ruo’ergai basins. Reflection geometries

76

X. Guo et al. / Earth and Planetary Science Letters 379 (2013) 72–80

(A)

(B) Fig. 4. (A) two close-up panels were selected to illustrate the seismic velocity structure beneath the Ruo’ergai basin and the Longriba fault zone; (B) two close-up panels were selected to present the crustal structure in the middle and southeastern portions of the seismic line (no vertical exaggeration and 6 km/s average seismic velocity is assumed); LRQF: Longriqu Fault; MJF: Minjiang Fault; WMF: Wenchuan-Maowen Fault; BCF: Beichuan Fault; PGF: Pengguan Fault.

in both the upper crust and at Moho indicate that the Yangtze crystalline crust does extend beneath the easternmost Tibetan Plateau to the LRQF, which is consistent with geochemical and geological studies (e.g. Burchfiel et al., 1995; Jiang and Jin, 2005; Roger et al., 2010). This continuous reflectivity across the Ruo’ergai basin to the LRQF clearly shows that the sedimentary cover was continuous while the crystalline basement was not. Thus, the thick Triassic sediments were deposited on a heterogeneous basement. The northwestern part of the LMS block contains distin-

guishable southeast-dipping reflectors, indicating backthrust deformation that separates variable-sized tilted crustal blocks. To the southeast, deep-seated, bidirectional-verging faults (southeast and northwest) indicate a pop-up structure is present that is a reasonable expectation gives the models of the LMS range front faulting of Hubbard et al. (2010). As two significant boundaries, both the LRQF of the Longriba fault zone and the WMF (Wenchuan-Maowen) of the LMS fault zone show strong crustal reflections that terminate at a depth co-

X. Guo et al. / Earth and Planetary Science Letters 379 (2013) 72–80

77

Fig. 5. The structural interpretation of the seismic profile (no vertical exaggeration and 6 km/s average seismic velocity is assumed); SGT: Songpan-Ganzi Terrane; LRQF: Longriqu Fault; MJF: Minjiang Fault; PGF: Pengguan Fault; WMF: Wenchuan-Maowen Fault; BCF: Beichuan Fault; PGF: Pengguan Fault.

inciding largely with the crust–mantle boundary (Moho). If lower crustal flow is an essential factor in the LMS block, horizontal reflectors would appear in the middle and lower crust, cutting through the Longriba and LMS fault zones, but they are not apparent. Most importantly, there is no observation that suggests lower crustal flow under the Ruo’ergai basin. Similarly, there is no evidence of a band of strong horizontal reflectors associating with the detachment fault in the upper crust to support the upper-crustal deformation model (Hubbard and Shaw, 2009). The mid-crust in other portions of the seismic reflection profile are acoustically transparent, and an explanation for this observation is not obvious. It is not a matter of signal strength since strong Moho reflections are observed. Currently, we attribute this scenario to the still active, and therefore, hot mid-crust of the LMS block, which is evident in the magnetotelluric studies of this area (Zhao et al., 2012). Regarding the structure of the mantle lithosphere, it is beyond the scope of our seismic reflection image so we don’t have specific information about it. However, we do believe shortening also occurred in the mantle lithosphere due to the complex tectonic setting of the research area. 5. Discussion We suggest that the GPS velocity gradient across the Longriba fault zone (Fig. 1a) is due to the existence of the highly shortened Yangtze crust beneath the LMS block (Fig. 5) that partly resisted the eastward movement of the Tibetan Plateau. This block is also associated with high topographic relief (Figs. 1c and 5) and the complex shortening structures east of the LMS range front (Hubbard and Shaw, 2009; Hubbard et al., 2010). Therefore, we propose that the highly shortened and thrusted LMS block is transferring stress from the Tibetan Plateau convergence to the LMS range front where it meets the rigid Sichuan basin (Fig. 5). In general, when the stress from eastward block motion of the Tibetan Plateau moved to the rear of the LMS block, elastic shortening began and caused the onset of stress transfer and diffusion (e.g., Bott and Dean, 1973) into the LMS block. A ‘stress front’ (Nemˇcok et al., 2005) accordingly advanced into the relatively unstressed region to the east. Given the relatively rigid nature of the intervening LMS block between the Longriba and the LMS fault zones, some stress is absorbed and generates intracrustal earthquakes with hypocenters deeper than 40 km (Fig. 6). This indicates the lower crust of the LMS block is capable of brittle failure as indicated in the seismic image. Meanwhile, some stress is transferred to the area beneath the LMS range and elastic shortening accumulates at depth. The major detachment imaged in the seismic data that connects the Wenchuan-Maowen, Beichuan, and Pengguan faults (panel IV in Fig. 4B) could be regarded as the principal channel transferring the stress to the range front. This interpretation is consistent with previous studies that ductile thickening within the deeper crust contributed to the LMS uplift (Royden et al., 1997;

Fig. 6. Seismicity in the Longmen Shan (LMS) fault zone area from 1959 to March 25, 2013 with magnitude 3.0. Dataset is derived from IRIS (URL: http://www.iris. edu/SeismiQuery/sq-events.htm). Colored dots represent seismic events with different hypocentral depths. Note that the earthquakes with hypocentral depths over 40 km are mainly distributed in the area between the Longmen Shan fault zone and the Longriba fault. Fault information is from Xu et al. (2008) and Nalbant and McCloskey (2011). LMS: Longmen Shan; MJ: Minjiang Fault.

Burchfiel et al., 2008). Release of the accumulated elastic shortening beneath the LMS range generates crustal-scale oblique extrusion and results in the uplift and right-lateral slip along the LMS. The extent of the LMS block beyond our seismic line can be estimated by the extent of the region of reduced GPS velocities (Fig. 1a) and the fault zones (Longriba and LMS) documented in the seismic image. Accordingly, we constructed a 2-D tectonic model that explains most of the diverse observations in the region. This model of the process of the eastward extrusion along the eastern edge of Tibet is illustrated in Fig. 7. In this cartoon, t1 refers to the time when a push is applied to the rear of the LMS block by the Tibetan Plateau; t2 is the time when the ‘stress front’ has moved to area beneath the LMS range and elastic shortening is accumulating; t3 refers to the time when the elastic shortening accumulation is released on the LMS fault zone during a large earthquake sequence and the next cycle of moment accumulation that could generate another serious earthquake begins. It is noteworthy that the recurrence interval of moment release illustrated in Fig. 7 is probably only one of several major earthquake sequences along the LMS fault zone. The 2008 Wenchuan earthquake is just the

78

X. Guo et al. / Earth and Planetary Science Letters 379 (2013) 72–80

Table 2 Deformation features of each tectonic unit between Mesozoic and Cenozoic. Timing

Tectonic units SGT

Western YB

LMS range

Early Mesozoic Late Triassic

Remnant Ocean(1) Compression and southeastward-directed thrust(4)

Passive margin(2) Inverted to active margin accompanied by offsets of Moho(5)

Late Mesozoic Timing after Late Quaternary(7)

Quiescence(6) Eastward extrusion in association with the block motion of the Tibetan Plateau(8)

Quiescence(6) Moho offsets are reactivated due to eastward extrusion of the Tibetan Plateau materials, and brittle failure in the lower crust occurred in a way of deep earthquake below 40 km depth(9)

Not yet initiated(3) Initiated as a sinistral thrust-Nappe belt during southeastward-directed thrusting toward the rigid SB(3) Quiescence(6) Uplift of the plateau margin occurred from 10–15 Ma to present(10) as a release of scalar moment that accumulated from eastward extrusion of Tibetan Plateau materials(9) ; sinistral slip movement was reversed(3)

SGT: Songpan-Ganzi Terrane; YB: Yangtze Block; SB: Sichuan Basin; LMS: Longmen Shan. References: (1) Yin and Nie (1993). (2) Harrowfield and Wilson (2005). (3) Burchfiel et al. (1995); Wang and Meng (2009). (4) Reid et al. (2005; 2007); Roger et al. (2011). (5) Roger et al. (2004; 2010). (6) Roger et al. (2011). (7) Xu et al. (2008). (8) Shen et al. (2005; 2009). (9) Study of this paper. (10) Wang et al. (2012).

only historically recorded one, so we use “e.g. the 2008 Wenchuan earthquake”. Thus our interpreted sequence of events is as follows. (a) As the oblique extrusion approached the western margin of the YB during the late Quaternary, dextral transpressive shearing of the Longriba fault zone was initiated as a result of resistance from the shortened crystalline crust to the east (Fig. 7t1 ). Stress from the block motion of Tibet is transferred within the highly shortened and thrusted crust of the LMS block up to the range front where it meets the relatively stable western Sichuan basin (part of the Yangtze hinterland) (Fig. 7t1 ). (b) As block motion continued, the ‘stress front’ advances within the fractured LMS block and elastic shortening gradually accumulates in the range front of the block (Fig. 7t2 ). During the moment accumulation, the LMS fault zone would have been relatively quiescent. (c) When the elastic shortening accumulated to a critical level, it would then be released suddenly during a large earthquake sequence (Zhu and Zhang, 2010), accommodating crustal-scale shortening and ductile thickening within the deeper crust. Release of the moment thus propagated upward along the Wenchuan-Maowen fault because of the resistance from the rigid Sichuan basement. Then it propagated along the major detachment fault to the Beichuan and Pengguan faults to the range front (Lu et al., 2012), resulting in the uplift of the LMS fault zone (Fig. 7t3 ). (d) With the release of the accumulated elastic shortening, the next cycle of moment accumulation would begin. The focus of the tragic Wenchuan earthquake and aftershocks are manifestations of this crustal-scale movement. This process also indicates that such destructive earthquakes will continue to occur in the LMS area as long as the oblique extrusion from the ongoing continent–continent collision continues. 6. Summary

Fig. 7. Interpreted tectonic evolution based on our interpretation of the deep seismic reflection profile and results of previous GPS studies (Chen and Wilson, 1996; Lv et al., 2003; Shen et al., 2005, 2009). Moho depth is obtained from the reflection times and the results of previous studies (Gao et al., 2005; Zhang et al., 2009; Wang et al., 2010). Shaded area of the Yangtze lower and middle crust at time t2 indicates the strain accumulation. A sequence of oblique ductile rectangular blocks represents the Yangtze crystalline crust that extends from the Sichuan basin to the Longriba fault area. SGT: Songpan-Ganzi Terrane; LMS: Longmen Shan; LRBFZ: Longriba Fault Zone; WMF: Wenchuan-Maowen Fault; BCF: Beichuan Fault; PGF: Pengguan Fault.

In summary, Table 2 is included to clarify the relationships between Mesozoic and Cenozoic deformation. More importantly, our 310 km-long, crustal-scale seismic reflection profile reveals detailed crustal-scale features in eastern Tibet for the first time and has provided strong evidence for the existence of Yangtze sub-continent crust beneath the easternmost edge of the Tibetan Plateau. Additionally, resulting models illustrate that the uplift of the LMS range is correlated with the release of deep-seated strain energy focused against the rigid Sichuan basin crust, which is consistent with the earthquakes occurring in the middle and lower crust (Fig. 6). Although other interpretations of some of the features we have discussed are certainly possible, the lack of subhorizontal zones of reflectivity within the crystalline do not sup-

X. Guo et al. / Earth and Planetary Science Letters 379 (2013) 72–80

port the channel flow (Royden et al. 1997, 2008) and upper-crustal deformation models (Hubbard and Shaw, 2009) in the LMS block area. We believe that our new observations and tectonic model advance our understanding of the tectonic responses of eastern Tibet to the ongoing India–Eurasia collision and the resulting natural hazards. Acknowledgements We thank all our colleagues from both China and USA who have worked in the field to collect the seismic data used in this study, and the No. 6 Geophysical Prospecting Team, Sinopec Huadong Petroleum Bureau, Nanjing, China, for their excellent and enthusiastic work under difficult terrain and logistic conditions. We would like to acknowledge B.C. Burchfiel, Guowei Zhang, An Yin, Mian Liu, Qiyuan Liu and Sheng Yu for their valuable comments on earlier versions of this manuscript and Chaoyang Kuang for numerous discussions on the processing flow of the seismic reflection data. This work was supported by the SinoProbe-02 Project, a National Natural Science Foundation of China grant (40830316) and a U.S. National Science Foundation, Partnerships for International Research and Education grant (0730154). Thanks to the two anonymous reviewers for their constructive comments and suggestions. Appendix A. Supplementary material Supplementary material related to this article can be found online at http://dx.doi.org/10.1016/j.epsl.2013.08.005. References Bott, M.H.P., Dean, D.S., 1973. Stress diffusion from plate boundaries. Nature 243, 339–341. Burchfiel, B.C., Chen, Z.L., Liu, Y.P., Royden, L.H., 1995. Tectonics of the LMS and adjacent regions, Central China. Int. Geol. Rev. 37, 661–735. Burchfiel, B.C., Royden, L.H., van der Hilst, R.D., Hager, B.H., Chen, Z., King, R.W., Li, C., Lü, J., Yao, H., Kirby, E., 2008. A geological and geophysical context for the Wenchuan earthquake of 12 May 2008, Sichuan, People’s Republic of China. GSA Today 18, 4–11. Cai, X.L., Wei, X.G., Liu, Y.C., Cao, J.M., 1996. On wedge-in orogeny–on the example of the Longmenshan orogenic belt. Acta Geologica Sichuan 16, 97–102 (in Chinese with English abstract). Chen, S.F., Wilson, C.J.L., 1996. Emplacement of the Longmen Shan thrust–Nappe belt along the eastern margin of the Tibetan Plateau. J. Struct. Geol. 18, 413–430. Chen, Z., Burchfiel, B.C., Liu, Y., 2000. Global Positioning System measurements from eastern Tibet and their implications for India/Eurasia intercontinental deformation. J. Geophys. Res. 105, 16215–16227. Clark, M.K., Bush, J.W.M., Royden, L.H., 2005. Dynamic topography produced by lower crustal flow against rheological strength heterogeneities bordering the Tibetan Pateau. Geophys. J. Int. 162, 575–590. Densmore, A.L., Ellis, M.A.Y., Li, Y., 2007. Active tectonics of the Beichuan and Pengguan faults at the eastern margin of the Tibetan Plateau. Tectonics 26, TC4005, http://dx.doi.org/10.1029/2006TC001987. Dong, S.W., Li, T.D., Gao, R., Hou, H.S., Li, Q.S., Li, Y.K., Zhang, S.H., Keller, G.R., Liu, M., 2011. A multidisciplinary Earth science research program in China. Eos, Trans. Am. Geophys. Union 92, 313–314. Gan, W.J., Zhang, P.Z., Shen, Z.K., Niu, Z.J., Wang, M., Wan, Y.G., Zhou, D.M., Cheng, J., 2007. Present-day crustal motion within the Tibetan Plateau inferred from GPS measurements. J. Geophys. Res. 112, B08416, http://dx.doi.org/ 10.1029/2005JB004120. Gao, R., Lu, Z.W., Li, Q.S., Guan, Y., Zhang, J.S., He, R.Z., Huang, L.Y., 2005. Geophysical survey and geodynamic study of crust and upper mantle in the Qinghai–Tibet Plateau. Episodes 28, 263–273. Harrowfield, M.J., Wilson, C.J.L., 2005. Indosinian deformation of the Songpan Garzê Fold Belt, northeast Tibetan Plateau. J. Struct. Geol. 27, 101–117. Hubbard, J., Shaw, J.H., 2009. Uplift of the LMS and Tibetan Plateau, and the 2008 Wenchuan (M = 7.9) earthquake. Nature 458, 194–197. Hubbard, J., Shaw, J.H., Klinger, Y., 2010. Structural setting of the 2008 Mw 7.9 Wenchuan, China, Earthquake. Bull. Seismol. Soc. Am. 100, 2713–2735. Jiang, X., Jin, Y., 2005. Mapping the deep lithospheric structure beneath the eastern margin of the Tibetan Plateau from gravity anomalies. J. Geophys. Res. 110, B07407, http://dx.doi.org/10.1029/2004JB003394.

79

Kirby, E., Harkins, N., Wang, E., Shi, X.H., Fan, C., Burbank, D., 2010. Slip rate gradients along the eastern Kunlun fault. Tectonics 26, TC2010, http://dx.doi.org/ 10.1029/2006TC002033. Lease, R.O., Burbank, D.W., Zhang, H.P., Liu, J.H., Yuan, D.Y., 2012. Cenozoic shortening budget for the northeastern edge of the Tibetan Plateau. Is lower crustal flow necessary? Tectonics 31, TC3011, http://dx.doi.org/10.1029/2011TC003066. Li, Y.Q., Jia, D., Shaw, J.H., Hubbard, J., Lin, A.M., Wang, M.M., Luo, L., Li, H.B., Wu, L., 2010. Structural interpretation of the coseismic faults of the Wenchuan earthquake. Three-dimensional modeling of the Longmen Shan fold-and-thrust belt. J. Geophys. Res. 115, B04317, http://dx.doi.org/10.1029/2009JB006824. Lu, R.Q., He, D.F., Suppe, J., Ma, Y.S., Liu, B., Chen, Y.Q., 2012. Along-strike variation of the frontal zone structural geometry of the Central Longmen Shan thrust belt revealed by seismic reflection profiles. Tectonophysics 580, 178–191. Lv, J.N., Shen, Z.K., Wang, M., 2003. Contemporary crustal deformation and active tectonic block model of the Sichuan-Yunnan region, China. Seismology and Geology 25, 543–554 (in Chinese with English abstract). Nalbant, S.S., McCloskey, J., 2011. Stress evolution before and after the 2008 Wenchuan, China earthquake. Earth Planet. Sci. Lett. 307, 222–232. Nemˇcok, M., Schamel, S., Gayer, R., 2005. Thrustbelts. Structural Architecture, Thermal Regimes, and Petroleum Systems. Cambridge University Press, New York. 541 pp. Nie, S., Yin, A., Rowley, D.B., Jin, Y., 1994. Exhumation of the Dabie Shan ultra-highpressure rocks and accumulation of the Songpan-Ganzi flysch sequence, central China. Geology 22, 999–1002. Reid, A.J., Wilson, C.J.L., Liu, S., 2005. Structural evidence for the Permo-Triassic tectonic evolution of the Yidun arc, eastern Tibetan Plateau. J. Struct. Geol. 27, 119–137. Reid, A.J., Wilson, C.J.L., Liu, S., Pearson, N., Belousova, E., 2007. Mesozoic plutons of the Yidun Arc, SW China. U/Pb geochronology and Hf isotopic signature. Ore Geol. Rev. 31, 88–106. Robert, A., Pubellier, M., de Sigoye, J., Vergne, J., Lahfid, A., Cattin, R., Findling, N., Zhu, J., 2010. Structural and thermal characters of the Longmen Shan (Sichuan, China). Tectonophysics 491, 165–173. Roger, F., Jolivet, M., Malavieille, J., 2010. The tectonic evolution of the SongpanGanze (north Tibet) and adjacent areas from Proterozoic to present. A synthesis. J. Asian Earth Sci. 39, 254–269. Roger, F., Jolivet, M., Cattin, R., Malavieille, J., 2011. Mesozoic-enozoic tectonothermal evolution of the eastern part of the Tibetan Plateau (Songpan-Ganze, Longmen Shan area). Insights from thermochronological data and simple thermal modeling. In: Gloaguen, R., Ratschbacher, L. (Eds.), Growth and Collapse of the Tibetan Plateau. In: Spec. Publ., vol. 353. Geological Society, London, pp. 9–25. Roger, F., Malavieille, J., Leloup, P.H., Calassou, S., Xu, Z., 2004. Timing of granite emplacement and cooling in the Songpan-Ganzi fold belt (eastern Tibetan Plateau) with tectonic implications. J. Asian Earth Sci. 22, 465–481. Royden, L.H., Burchfiel, B.C., van der Hilst, R.D., 2008. The geological evolution of the Tibetan Plateau. Science 321, 1054–1058. Royden, L.H., Burchfiel, B.C., King, R.W., Wang, E.C., Chen, Z.L., Shen, F., Liu, Y.P., 1997. Surface deformation and lower crustal flow in eastern Tibet. Science 276, 788–790. Shen, Z.K., Lü, J.N., Wang, M., Bürgmann, R., 2005. Contemporary crustal deformation around the southeast borderland of the Tibetan Plateau. J. Geophys. Res. 110, B11409, http://dx.doi.org/10.1029/2004JB003421. Shen, Z.K., Sun, J.B., Zhang, P.Z., Wan, Y.G., Wang, M., Burgmann, R., Zeng, Y.H., Gan, W.J., Liao, H., Wang, Q.L., 2009. Slip maxima at fault junctions and rupturing of barriers during the 2008 Wenchuan earthquake. Nat. Geosci. 2, 718–724. Wang, C.S., Gao, R., Yin, A., Wang, H.Y., 2011. A mid-crustal strain-transfer model for continental deformation: a new perspective from high-resolution deep seismicreflection profiling across NE Tibet. Earth Planet. Sci. Lett. 306, 279–288. Wang, C.Y., Zhu, L.P., Lou, H., Huang, B.S., Yao, Z.X., Luo, X.H., 2010. Crustal thicknesses and Poisson’s ratios in the eastern Tibetan Plateau and their tectonic implications. J. Geophys. Res. 115, B11301, http://dx.doi.org/ 10.1029/2010JB007527. Wang, E., Kirby, E., Furlong, K.P., van Soest, M., Xu, G., Shi, X., Kamp, P.J.J., Hodges, K.V., 2012. Two-phase growth of high topography in eastern Tibet during the Cenozoic. Nat. Geosci, http://dx.doi.org/10.1038/NGEO1538. Wang, E.C., Meng, Q.R., 2009. Mesozoic and Cenozoic tectonic evolution of the Longmenshan fault belt. Sci. China D. Earth Sci. 52, 579–592. Xu, X.W., Wen, X.Z., Chen, G.H., Yu, G.H., 2008. Discovery of the Longriba Fault zone in eastern Bayan Har Block, China and its tectonic implication. Sci. China D. Earth Sci. 51, 1209–1223. Yin, A., 2010. A special issue on the great 12 May 2008 Wenchuan earthquake (Mw 7.9): Observations and unanswered questions. Tectonophysics 491, 1–9. Yin, A., Nie, S., 1993. An indentation model for North and South China collision and the development of the Tanlu and Honam fault systems, eastern Asia. Tectonics 12, 801–813. Zhang, G.W., Guo, A.L., Yao, A.P., 2004. Western Qinling-Songpan continental tectonic node in China’s continental tectonics. Earth Sci. Front. 11, 23–32 (in Chinese with English abstract). Zhang, H.F., Parrish, R., Zhang, L., Xu, W.C., Yuan, H.L., Gao, S., Crowley, Q.G., 2007. A-type granite and adakitic magmatism association in Songpan-Garze fold belt,

80

X. Guo et al. / Earth and Planetary Science Letters 379 (2013) 72–80

eastern Tibetan Plateau. Implications for lithospheric delamination. Lithos 97, 323–335. Zhang, H.F., Zhang, L., Harris, N., Jin, L.L., Yuan, H.L., 2006. U–Pb zircon ages, geochemical and isotopic compositions of granitoids in Songpan-Ganze fold belt, eastern Tibetan Plateau. constraints on petrogenesis and tectonic evolution of the basement. Contrib. Mineral. Petrol. 152, 75–88. Zhang, Z.J., Wang, Y.H., Chen, Y., Houseman, G.A., Tian, X.B., Wang, E.C., Teng, J.W., 2009. Crustal structures across Longmenshan fault belt from passive source seismic profiling. Geophys. Res. Lett. 36, L17310, http://dx.doi.org/ 10.1029/2009GL039580.

Zhao, G.Z., et al., 2012. Crustal structure and rheology of the Longmenshan and Wenchuan Mw 7.9 earthquake epicentral area from magnetotelluric data. Geology 40, 1139–1142. Zhou, D., Graham, S.A., 1996. Songpan-Ganzi Triassic flysch complex of the West Qinling Shan as a remnant ocean basin. In: Yin, A., Harrison, M. (Eds.), The Tectonic Evolution of Asia. Cambridge University Press, Cambridge, pp. 281–299. Zhu, S.B., Zhang, P.Z., 2010. Numeric modeling of the strain accumulation and release of the 2008 Wenchuan, Sichuan, China, earthquake. Bull. Seismol. Soc. Am. 100, 2825–2839.