Crustal structure and deformation under the Longmenshan and its surroundings revealed by receiver function data

Physics of the Earth and Planetary Interiors 244 (2015) 11–22

Contents lists available at ScienceDirect

Physics of the Earth and Planetary Interiors journal homepage: www.elsevier.com/locate/pepi

Crustal structure and deformation under the Longmenshan and its surroundings revealed by receiver function data Ya Sun a,b,c,⇑, Jianxin Liu a,b, Keping Zhou c, Bo Chen a,b,⇑, Rongwen Guo a,b a

School of Geosciences and Info-Physics, Central South University, Changsha, China Key Laboratory of metallogenic prediction of Non-Ferrous metals (Central South University), Ministry of Education, Changsha, China c School of Resources and Safety Engineering, Central South University, Changsha, China b

a r t i c l e

i n f o

Article history: Received 27 November 2014 Received in revised form 23 April 2015 Accepted 23 April 2015 Available online 5 May 2015 Keywords: Longmenshan region Crustal seismic anisotropy Lower crust flow Upward extrusion

a b s t r a c t The convergence of India and Eurasia and the obstruction from the rigid Sichuan Basin cause the Longmenshan (LMS) to have the steepest topographic gradient at the eastern margin of the Tibetan Plateau. However, the mechanisms of surface uplift are still controversial. In this paper, we estimate the crustal structure and deformation under the LMS and its surroundings by analyzing a large amount of receiver function data recorded by regional seismic networks of the China Earthquake Administration. We apply a comprehensive splitting measurement technique on Ps conversion phase at the Moho (Moho Ps splitting) to calculate crustal anisotropy from azimuthal variations of receiver functions. Our results show that most of the seismic stations beneath the LMS area exhibit significant seismic anisotropy with the splitting time of 0.22–0.94 s and a fast polarization direction of NW–SE, while less or even no crustal anisotropy has been observed under the Sichuan Basin. Comparing the fast polarization directions of Moho Ps splitting with the indicators of lithospheric deformation (such as shear wave splitting, absolute plate motion, and global positioning system) imply a consistent tendency of deformation between the lower crust and upper mantle, but decoupling deformation in the crust beneath the LMS area. We further compare Moho Ps splitting time to that estimated from previous SKS splitting, indicating that crustal anisotropy is an important source of the SKS splitting time in this study area. In addition, a thick crust (>50 km) with high Vp/Vs values (1.74–1.86) is also observed using the H-j stacking method. These seismic observations are consistent with the scenario that the LMS area has been built by the lower crustal flow. Combined with the seismic reflection/refraction profile and geology studies, we further suggest that the lower crustal flow may extrude upward into the upper crust along the steeply dipping strike faults under the LMS area, resulting in the surface uplift of the LMS. Crown Copyright Ó 2015 Published by Elsevier B.V. All rights reserved.

1. Introduction The continental collision between India and Eurasia in the Cenozoic has resulted in significant crustal shortening across Asia and uplifting of the Tibetan Plateau (Molnar and Tapponnier, 1975; Houseman and England, 1993; Clark and Royden, 2000). Due to the influences of the northward subduction of the Indian plate (Ni et al., 1989) and the resistance of the rigid Sichuan Basin (Clark et al., 2005), the Longmenshan (LMS) belt as the eastern boundary of Tibetan Plateau is dominated by strong compressional deformation and active intra-continent seismicity. The LMS front exhibits sharp changes both in the crustal thickness (from ⇑ Corresponding authors at: School of Geosciences and Info-Physics, Central South University, Changsha, China. Tel: +86 0731 88876939. E-mail addresses: [email protected] (Y. Sun), [email protected] (B. Chen). http://dx.doi.org/10.1016/j.pepi.2015.04.005 0031-9201/Crown Copyright Ó 2015 Published by Elsevier B.V. All rights reserved.

40 km in the east to 60–65 km in the west) (Zhang et al., 2009; Xu et al., 2007) and in topographic elevation (from 500 m above sea level in the Sichuan Basin to 4–6 km on the eastern Tibetan Plateau) without significant the upper crustal shortening observed from the absolute plate motion (GPS) data (<3 mm/yr) (Burchfiel et al., 2008; Shen et al., 2005). Two end-member models have been developed to describe the tectonic evolution and surface uplift of the LMS, such as lateral extrusion of rigid blocks (Tapponnier et al., 1982) and lower crustal flow (Royden et al., 1997; Clark and Royden, 2000). The two models represent fundamentally different views on the deformation mechanism of the Plateau, and imply different dynamic processes in the crust and mantle. For example, the rigid block model requires strong lithospheric blocks with strain localized along major shear zones (Tapponnier et al., 1982). On the other hand, lower crustal flow model explains the lateral expansion of the Plateau by an injection of ductile lower

12

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

crustal material from the central Tibetan Plateau, which implies mechanical decoupling between the upper crust and upper mantle. Determining the depth variation in the lithospheric deformation thus is of great importance for better understanding the orogenic process. As a proxy for deformation, seismic anisotropy plays an important role in constraining the mode and location of the Earth’s deformation (e.g., Nicolas and Christensen, 1987; Mainprice and Nicolas, 1989). In general, seismic anisotropy in the Earth’s upper crust is caused by stress-induced alignment of cracks, while it in the lower crust and mantle is usually attributed to strain-induced lattice-preferred orientation of the minerals in the crust and mantle. For example, if lower crustal flow is the dominant mechanism for the observed surface uplift of the LMS, the fast direction of crustal anisotropy may align to the flow direction, due to the lattice preferred orientation through deformation (Silver, 1996; Sun et al., 2012). Thus seismic anisotropy has been widely used to quantify the deformation associated with a range of tectonic processes (McNamara and Owens, 1993; Silver, 1996). Shear wave splitting (SKS and Ps conversion wave at the Moho) and surface wave dispersion data have been used to measure the seismic anisotropy to investigate subsurface deformation in the LMS area (Wang et al., 2008; Chang et al., 2008; Yao et al., 2010; Chen et al., 2013). SKS splitting data combined with GPS measurements suggested the vertical coherent deformation between the crust and upper mantle beneath most of southeastern Tibet (Meltzer et al., 2007; Wang et al., 2008; Chang et al., 2008). However, Yao et al. (2010) found that the crust has some azimuthal anisotropy with a fast polarization direction different from that of the mantle in southwest of the LMS belt, indicating a crust-mantle decoupling deformation. Moreover, Chen et al. (2013) found crustal anisotropy with splitting times ranged from 0.1 to 0.26 s near the LMS area using Pms converted waves at the Moho from individual receiver function data, indicating that crustal anisotropy could not be ignored. In general, GPS measurements reflect the deformation of the brittle upper crust, while SKS is a P-to-S converted wave at the core-mantle boundary and generally has a poor resolution on the source depth of the measured vertical extent of seismic anisotropy (Lev et al., 2006). Pms converted waves at the Moho are relatively weaker compared to the SKS/SKKS phase. It is extremely difficult to obtain robust measurements of crustal anisotropy from individual receiver function data (Liu and Niu, 2012). Recently, Liu and Niu (2012) developed a splitting measurement technique based on Ps converted waves at the Moho from the receiver functions (Moho Ps splitting) in the whole range of back azimuths to measure azimuthal S-wave anisotropy with a horizontal axis in crust. Based on this method, Sun et al. (2012) added a harmonic analysis component to isolate the azimuthal S-wave anisotropy with a horizontal axis from other heterogeneous structures (e.g., dipping Moho, P wave anisotropy, and azimuthal S-wave anisotropy with an inclined axis) and applied them on southeastern Tibet to study the crustal deformation. In this paper, we apply this improved method to the new broadband seismic data recorded by regional seismic networks operated by the China Earthquake Administration (CEA) to study the crustal azimuthal S-wave anisotropy with a horizontal axis in the LMS and its surroundings. For brevity, crustal anisotropy in the following sections represents azimuthal S-wave anisotropy with a horizontal axis in crust. We chose a total of 58 stations to measure crustal anisotropy, Moho depth and Vp/Vs ratio using receiver function data. Subsequently, our results are compared with those estimated from GPS, SKS and absolute plate motion (APM) to constrain the deformation process of the crust and upper mantle, and to shed light on the dynamics of the surface uplift beneath the LMS and its surroundings.

2. Data and analysis 2.1. CEA regional network data From 2003 to 2009, the CEA constructed the Chinese digital earthquake observation network, which aims to improve the earthquake monitoring and to provide data for seismologists and geodynamicists to study the fine structures and dynamics. This project includes about 1000 digital seismic stations, 32 regional seismic networks, 6 seismic arrays and 6 volcano monitoring networks (Zheng et al., 2010). All stations are equipped with a 24-bit digitizer. To study lateral variations of the crustal structure and seismic anisotropy beneath the LMS area, we selected 58 stations located in the region between 98° and 106° in longitude, and 28– 34° in latitude (Fig. 1b). Most of stations are recorded at a sampling rate of 100 Hz, with some exceptions of 50 Hz in the broadband seismometers. The instrument response of the stations with the seismometers are flat up to about 360 s, 120 s and 60 s, respectively, with a zero phase response in the flat band (Zheng et al., 2010). In this study, we use the sampling rate of 100 Hz in the receiver function data to measure crustal anisotropy. We visually examined a total of 413 earthquakes recorded from July, 2007 to July, 2010, with magnitudes greater than 5.0, and located within an epicentral distance of 30–90° for each station. These earthquakes provide a good coverage in back azimuth (Fig. 2). 2.2. Receiver function and harmonic analysis of Ps arrival time To generate receiver functions, we first rotated the two horizontal components into the radial (R) and transverse (T) components with sensor orientation corrected at each station using the method of Niu and Li (2011), and then rotate the radial (R) and vertical (V) components into the P and SV coordinates (Niu and Kawakatsu, 1998). We use the ‘‘water-level’’ deconvolution technique to build R, T, and SV receiver functions after the coordinate rotation (Clayton and Wiggins, 1976; Ammon, 1991). We set the ‘‘water-level’’ and the corner frequency of the Gaussian low-pass filter here to be 0.01 and 4.0, respectively, which is equivalent to a corner frequency of 0.5 Hz. Then, we normalize the R and T receiver functions in the time window between 5 s and 35 s. The R and T receiver functions are used in the harmonic analysis and estimation of seismic anisotropy; and SV receiver functions are used to calculate Moho depth and Vp/Vs ratio in the H-j analysis. To eliminate noisy events, we visually examine all the receiver functions and remove those with low signal-to-noise ratio (SNR < 0.85). Further, we calculate the covariance matrix of all the receiver functions at each station and eliminate the ones that showed a low cross correlation coefficient (<0.7) (Chen et al., 2010). Finally, the number of receiver functions range from 15 to 303 with an average of 198 are obtained at 58 stations (Table 1S). In general, the relative arrival times of Moho Ps (P-to-S converted phase at the Moho) have negative distance moveouts due to negative ray parameters with respect to the direct P wave. According to the revised IASP91 velocity model (Kennett and Engdahl, 1991), we compute the moveouts and make corrections, thereby all the Moho Ps phases have a relative arrival time referred to an arrival with an epicentral distance of 60° and a source depth of 0 km. After the moveout correction and normalization, the R and T receiver functions are shown as a function of back azimuth to detect systematic variations in the peak Ps arrival time and polarity changes in the R and T receiver functions. To evaluate the measurement reliability of the seismic anisotropy (see the following section), we first apply the harmonic analysis on Moho Ps arrival times before estimating crustal anisotropy (Sun et al., 2012). Because different heterogeneous structure in

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

13

Fig. 1. (a) Maps show the surface motions of Indian plate and different blocks within the Tibetan Plateau relative to the Eurasia. The black box indicates the study region. (b) Topographic map shows the CEA broadband station (solid triangles); yellow lines indicate the major faults in our study area. Dash black line in (b) is the location of seismic refraction profile in Chen et al. (2013), while a black line aligns the dash line represents the location of the cartoon profile shown in Fig. 9. Abbreviations are: LMSF, Longmenshan Fault; SGB, Songpan–Ganzi Block; XSHF, Xianshuihe Fault; JRS, Jinsha River Suture; CDB, Chuan–Dian Block; LTF, Litang Fault. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

crust can cause a harmonic moveout in Moho Ps arrival time with the azimuthal, it might confuse the features those are used to calculate crustal anisotropy (McNamara and Owens, 1993; Liu and Niu, 2012). Assuming a harmonic degree n with peak-to-peak amplitude of dt, and initial phase of u, the harmonic moveout correction at a station with a back azimuth of hi can be written as (Sun et al., 2012):

dt i ¼

dt cos ðnhi þ uÞ 2

ð1Þ

We correct the harmonic moveout dti for the R receiver functions in a time window length of tL (3 s) centered on t0 to find systematic variations in the arrival time of Moho Ps as a function of back azimuth. We then stack all the R receiver functions:

F r ðt; u; dtÞ ¼

N X F ir ðt  dti Þ; t ½t 0  0:5tL ; t 0 þ 0:5tL 

ð2Þ

i¼1

where superscript i represents the i-th receiver function and N is the total number of the receiver functions. t0 is the average arrival time of Moho Ps phase, and it can be obtained by stacking all the R receiver functions (Fig. 3a and b). The normalized maximum amplitude (An,max), maximum energy of the stacked receiver function (En,max), and minimum total residual (Rn,min) between each receiver function and the stacked receiver function can be calculated. We varied n from 1 to 8, hi in the range of 0–360° with an increment of 1° and dt from 0.0 to 1.5 s with an increment of 0.02 s. An example of the harmonic analysis at station MXI is shown in Fig. 3c.

14

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

0˚

0˚

60

0

2

4

6

8

10

12

14

-2

0

2

4

6

8

10

12

14

0˚

0˚

100 150

0˚

12

0˚

24 21

Delta (deg)

90˚

270˚

˚

30

-2

50

˚

33

SC.MXI BHR

a

30

0˚

Geophysical studies (e.g., Levin and Park, 1997; Peng and Humphreys, 1997; Shiomi and Park, 2008) showed that crustal anisotropy can generate a degree-2 azimuthal variation in Moho Ps arrival times, while several other types of heterogeneous crustal structures (e.g., dipping Moho, P-wave anisotropy or azimuthal S-wave anisotropy with an inclined axis) can result in a degree-1 back azimuthal variation, and small-scale azimuthal variations in crustal velocity and Moho topography may result in higher order harmonic variations in the Moho Ps arrival time. Thus, a distinct peak of the maximum amplitude and energy of the stacked receiver function, as well as the best fit between the stacked and individual receiver functions at degree-2 can be considered as a manifestation of crustal anisotropy (Fig. 3c). 2.3. Crustal anisotropy estimation

250

Fig. 2. We use all the seismograms with an epicentral distance between 30° and 90°. A total of 413-teleseismic events (red solid circles) are used in this study. Most of the earthquakes are located in the western and south Pacific, as well as the Indonesia subduction zone. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

300

180˚

200

15

b

T0 -2

c

0

2

4

6

8

10

1.50

14

Аn,max En,max 1/Rn,min

1.40

When radially polarized S waves propagate through an anisotropic layer with a horizontal symmetrical axis, the S wave is split into two distinct quasi-S waves with different polarizations and velocities. One quasi-S wave will be projected onto the transverse component. The arrival time of the Moho Ps converted waves on the R and T components and the polarity of S waves on the T components each exhibits a four-lobe (harmonic degree-2) variation pattern along the back azimuth direction (Sun et al., 2012). In addition, the birefringent T waveforms are proportional to the time derivative of those on the R component. These two unique features provide strong indicators for resolving crustal anisotropy from other heterogeneous structures, such as velocity heterogeneities in the crust and a dipping Moho. The method developed by Liu and Niu (2012) are used to constrain crustal anisotropy in this paper based on the two unique features described above. This method includes computing crustal anisotropy using three individual and one joint objective functions,

12

1.30 1.20 1.10 1.00

1

2

3

4

5

6

7

8

Fig. 3. (a) An example of SV receiver functions from station MXI are plotted as a function of back azimuth. The receiver functions here are stacked in 10° bins along the back azimuth direction. The solid line indicates the average arrival times of Moho Ps conversion phase, and the brown circles represent the azimuthal variation in its arrival time. (b) The 1nd-root stacked receiver functions. P to S conversions at the base of the sediment layer and the crust are clearly shown. T0 represent the average of Ps arrival time. (c) The harmonic analysis at the same station. The maximum value of peak amplitude and total energy, as well as the reciprocal of the minimum residual are shown at the harmonic degree-2, indicating azimuthal seismic anisotropy with a horizontal axis at this station. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

and a statistical analysis of the reliability of the estimated anisotropy with the T and R receiver functions. We apply this method in the time window of 2–8 s in the R and T receiver functions. The individual objective functions are designed to search for crustal anisotropy parameters (the fast direction (u) and the splitting time (dt) in the (u, dt) domain) (Liu and Niu, 2012) via the following operations on Ps arrival times: (1) maximizing the peak energy of the R receiver function stacked after a cosine correction (Fig. 4a); (2) maximizing the cross-correlation of the radial receiver functions after the removing of crustal anisotropy (Fig. 4b); and (3) minimizing the total energy of the T receiver functions stacked after removing crustal anisotropy (Fig. 4c). The joint objective function is calculated by the weighted average of three individual objective functions (Fig. 4d), while the statistical analysis of stacking a total of N coherent signals can result in an increased of SNR by a factor of N1/2 (Fig. 4e and f). An example of the results at station MXI is shown in Fig. 4. As suggested by Sun et al. (2012) and Liu and Niu (2012), if anisotropy is significant, the SNR of the T receiver functions stacked after correction of waveform polarization (open squares in Fig. 4e) and the R receiver functions stacked after cosine correction of anisotropy (filled triangle in Fig. 4f) across the whole range of N1/2 are expected to be larger than those uncorrected (filled squares and open triangles in Fig. 4e and f, respectively). Once crustal anisotropy is corrected, only random noise remains in the T component, and the SNR of the T receiver functions will be flat across the whole range of N1/2, regardless of whether the polarity correction is applied (filled circles in Fig. 4e) or not (open circles in Fig. 4e). Parts of SV and T receiver functions before and after the correction of crustal anisotropy at station MXI are shown in Fig. 5a and b, respectively. The arrival time of Moho Ps can be easily observed with clear azimuthal variations. Once seismic anisotropy is corrected, these features are no longer seen from the data (Fig. 5c and d). Although we have isolated crustal anisotropy from other heterogeneous structures by the above harmonic analysis, we further separate crustal anisotropy and a dipping Moho (a distinct feature beneath the LMS belt) using the amplitude variation shown on the joint objective function images of measuring crustal anisotropy. Liu and Niu (2012) performed the synthetic test of a dipping Moho model and found the low maximum amplitude (less than 1.1) and small splitting time (less than 0.1 s) shown on the joint objective image of measuring crustal anisotropy. Thus, the large splitting time (0.52 s) and maximum amplitude (1.907) of the joint objective function image (Fig. 4) manifest a significant crustal anisotropy beneath station MXI, which is not caused by a dipping Moho, while a low maximum amplitude (1.013) and small splitting time (0.08) shown on the joint objective image of measuring crustal anisotropy suggest insignificant crustal anisotropy beneath station ZJG (Fig. 1S). 2.4. H-j stacking analysis The crustal thickness and Vp/Vs ratio of the crust are calculated at each station using the H-j stacking method (Niu et al., 2007; Chen et al., 2010), which is improved from the method proposed by Zhu and Kanamori (2000). This improved H-j search technique can introduce a large trade off between H and j using a coherence index of the three phases, c(j):

( sðH; jÞ ¼

) N cðjÞ X ðw1 ri ðt1 Þ þ w2 ri ðt2 Þ  w3 r i ðt3 ÞÞ N i¼1

ð3Þ

where N is the number of receiver functions at a given station and ri(t) represents the amplitude of the ith receiver function at the predicted arrival times of the primary P-to-S converted phase 0p1s (t1),

15

and two Moho reverberation phases, 2p1s (t2) and 1p2s (t3) (Niu and James, 2002; Niu et al., 2007). w1, w2 and w3 are the weights of the three phases. We first calculate the initial crustal thickness using a first-root stacking technique (Kawakatsu and Niu, 1994) by assuming that P-to-S conversions are the primary sources of energy in the P-code window. We search for the conversion depth in the range of 0–80 km with an increment of 1 km and consider the depth with the maximum amplitude as the initial crustal thickness. The stacked depth profile at station MXI (Fig. 6a) shows a clear P-to-S conversion peak at depth of 46.5 km. Further, the weights of 0.5, 0.25, 0.25 for the three phases 0p1s, 2p1s and 1p2s (Niu et al., 2007) are employed to calculate the crustal thickness (H) and average Vp/Vs ratios (j) in crust. We search for H in the range of 20–80 km with an increment of 1 km. j is varied in the range of 1.55–2.0 with an increment of 0.001. The H and j ratios are finally determined by picking the location where the summed amplitude s(H, j) reaches its maximum. As shown in Fig. 6b, a well-defined peak at (H = 45.3 m, Vp/Vs = 1.761) is estimated on the basis of the corrected crustal anisotropy in receiver function data. 3. Results Finally, we obtain a subset of 21 measurements for crustal anisotropy from the 58 stations, which are listed in Table 1S and shown in Fig. 7. For a weakly anisotropic or an isotropic medium under the stations, the Ps arrival times do not show a harmonic degree-2 azimuthal variation, and the difference between the receiver functions before and after the correction of seismic anisotropy is not corresponding to the SNR test of Fig. 4e and f. Both of them are expected to be insignificant and not shown in crustal anisotropy images (Fig. 7a and b). Most of these 21 measurements are geographically confined to the west of the LMS belt, and the measured splitting time varies from 0.22 s to 0.94 s with an average of 0.57 s. The fast polarization directions beneath most of these measurements are NW-SE and nearly perpendicular to the LMS fault, which are consistent with the azimuthal anisotropy observed from P wave data inside the crust of the LMS area (Wei et al., 2013). The fast polarization directions are also in good agreement with those measured from the Moho Pms shear wave splitting from individual receiver functions (Chen et al., 2013), while the splitting times are much larger than those measured from the latter. This may because the Moho Pms conversion wave from individual receiver functions (Chen et al., 2013) is a much weaker signal, which provides a reliable splitting measurement only when noise levels in the data are lower than 30% (Liu and Niu, 2012). However, both harmonic and statistical analyses have been applied on the stacking receiver function data with a good azimuthal coverage in this study to ensure that the estimated splitting parameters are not artifacts. We also obtain subsets of 51 measurements of crustal thickness (H) and of 43 measurements of Vp/Vs ratios (j) from the 58 stations using the H-j stacking method. The Moho depth from the sea level (D) is further computed by subtracting station elevations from the measured H. Both Moho depth and Vp/Vs ratio are interpolated into meshed 0.2°  0.2° grids of the study area (Fig. 8a and b). The estimated Moho depth varies from 38.6 km in the Sichuan Basin to 67.3 km in southeastern Songpan–Ganzi block. The high Vp/Vs ratio (1.74–1.86) is distributed under the LMS and its surroundings, and decreases across the northwest part of the Sichuan Basin to less than 1.70. The Moho depth and Vp/Vs ratio are roughly consistent with the results of Xu et al. (2007) and Zhang et al. (2009). Since there are few stations in the northwest and southwest corners of the study area shown in Fig. 8, the Moho depth and Vp/Vs ratio for these two corners are not well constrained.

16

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

Fig. 4. Results obtained from the joint analysis of crustal anisotropy at station MXI. a, b and c correspond to three individual methods of estimating seismic anisotropy as shown in the paper. The joint function of estimating seismic anisotropy is shown in d. Note that the circles in every image of estimating seismic anisotropy represent the splitting time. Open and filled symbols in e and f represent SNR calculated from stacks of receiver functions before and after the removal of seismic anisotropy determined by the joint receiver function. Note the steady increase of the open squares with increasing N1/2 in e and that the filled triangles are always above the open ones in f.

4. Discussion 4.1. Moho depth and Vp/Vs ratio As shown in Fig. 8a, the crustal thickness deepens gradually toward the northwest from the eastern Sichuan Basin (38 km), and reaches 50 km in the LMS belt. Further west, the crust thickens to more than 60 km in the southeastern Songpan–Ganzi block. Vp/Vs ratio is high (1.74–1.86) in the LMS and southeastern Songpan–Ganzi block and low (1.70) in the Sichuan Basin (Fig. 8b). In general, Vp/Vs ratio in the crust contains information about crustal mineralogy, and can be used to constrain its bulk average composition (Christensen, 1996). Mineral physics studies (Christensen, 1996; Owens and Zandt, 1997) indicated that Vp/Vs ratio increases with an increase in plagioclase content or a

decrease in quartz content of a rock, and high Vp/Vs ratios (>1.87) may indicate the presence of partial melt. Fluids in the lower crust can decrease the average crustal shear wave velocity and increase Vp/Vs ratio (Makovsky et al., 1996). A receiver function study (Zhang et al., 2009) suggested that high Vp/Vs ratio in Songpan–Ganzi and the LMS belt (compared to global averages (Christensen, 1996)) may relate to the overall mafic composition of the crust, and/or the existence of fluids in the lower crust. Mineral analyses (Xu et al., 2001; Zhang et al., 2006) suggested that the mafic–acidic volcanic rocks erupted from lower crust or lithospheric mantle are widely distributed in eastern Tibet after India– Asia collision. All those imply that the observed high Vp/Vs ratio (1.74–1.86) under the LMS area may relate to the fluids or mafic materials, because mafic usually contains gabbro and peridotite or dunite (Gao et al., 1998; Sun et al., 2012).

17

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

MXI T RFs before correction

SV RFs before correction 0

2

-2

0

2

4

6

8

10

12

14

4

6

8

10

12

14

b

-2

0

2

-2

0

2

4

6

8

10

12

14

4

6

8

10

12

14

100

Delta (deg)

120

100

180

180

160

160

140

140

120

Delta (deg)

80

80

60

60

40

40

a

-2

Time (sec)

Time (sec)

SV RFs after correction 0

2

-2

0

2

4

6

8

10

12

14

4

6

8

10

12

14

d

T RFs after correction -2

0

2

-2

0

2

4

6

8

10

12

14

4

6

8

10

12

14

40

-2

180

180

160

160

140

140

120

100

Delta (deg)

100 120

Delta (deg)

80

80

60

60

40

c

Time (sec)

Time (sec)

Fig. 5. Part of R and T receiver functions shown here are recorded at station MXI by binned in 10° azimuthal caps (a–d). A comparison between receiver functions before (a and b) and after the correction (c and d) for seismic anisotropy. The solid line indicates the average arrival time of the Moho Ps converted phase. Brown circles in (a) indicate the arrival time of the peak amplitude of the Ps converted phase at Moho with a cosine variation along the back azimuthal direction. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

18

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

a

20

0

b

40 Depth (km)

60

In the Sichuan Basin, Fig. 8b shows that Vp/Vs ratios are over a large range (1.68–1.80). We attribute this variation to the changes in sediment thickness within the basin (Watson et al., 1987; Yong et al., 2003; Sun et al., 2012). Because Vp/Vs ratio is a weighted average of the rocks in a crustal column and sediment, and the sediment commonly has a high Vp/Vs ratio, the higher proportion of sediment in the crustal column may result in a higher Vp/Vs ratio. In addition, the strong sediment reverberations observed between the Moho Ps and the direct P wave in the receiver functions beneath most stations in the Sichuan Basin (e.g., station MXI in Fig. 3a) manifest that the sedimentary-bedrock boundary is located at a few to ten kilometers depth from the surface, which may be the dominant factor in the obvious variation of Vp/Vs ratio in the Sichuan Basin.

80

2.00

Vp/Vs

1.95 1.90

4.2. Primary cause of crustal anisotropy (Moho Ps splitting)

1.85

Since the Moho Ps conversion phase is confined in the crust, the splitting time of Moho Ps is a weighted average of the whole crust. Many studies (McNamara and Owens, 1993; Crampin and Peacock, 2008) indicated that shear wave splitting in the crust is mainly caused by micro-cracks in the upper crust, and the splitting time is generally less than 0.2 s in the LMS area (Shi et al., 2013). Comparing the observed significant average splitting time of Ps (0.57 s), we can then deduce that the significant splitting time may be contributed to the middle-to-lower crust. In addition, in the southeastern Tibetan Plateau, Sun et al. (2012) showed that an anisotropic model of 6% azimuthal anisotropy in the lower crust (0.0165 s/km) may be caused by the lattice preferred orientation of amphibole associated with the lower crustal flow. Since the observed average Moho depth under the LMS is 50 km, given an average upper crust thickness of 15 km (Xu et al., 2008; Jia et al., 2014), the average lower crust thickness is 35 km. According to the anisotropic model (Sun et al., 2012), we obtain an average splitting time of 0.58 s in the lower crust, which is close to the measurements with the average splitting time of 0.57 s in this study. Therefore, seismic anisotropy observed by

1.80 1.75 1.70 1.65 1.60 1.55 30

40

50 Depth (km)

60

70

Fig. 6. (a) The stacked receiver functions after the time to depth conversion at station MXI. The first peak located between the direct P wave and Moho Ps conversion phase is associated with the sediment–basement boundary. The second peak of the Ps conversion from Moho is at 46.5 km. (b) H-j analysis result of the receiver functions with correction of crustal anisotropy. The estimated Moho depth is 45.3 km and Vp/Vs ratio is 1.761.

34˚

34˚

a

32˚

32˚

30˚

30˚

b

C

28˚ 98˚

100˚ Ps

102˚ 0.4s

GPS

104˚ 12mm/yr

106˚ Pms

28˚ 98˚ 0.2s

100˚ SKS

0.4s

102˚ AMP

104˚

106˚

18mm/yr

Fig. 7. Comparison of crustal anisotropy from Ps phase splitting analysis (blue bar line with a circle) with (a) crustal anisotropy measured from Pms (Chen et al., 2013) (yellow arrows) and GPS velocities relative to the South China Block (Shen et al., 2005; Burchfiel et al., 2008) (red arrows), and (b) SKS/SKKS splitting time (Wang et al., 2008; Chang et al., 2008) (black bars) and the direction of absolute plate motion (APM) with respect to GSRM V1.2 models (Kreemer et al., 2003) (thick white arrows with black frames). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

Moho Ps splitting from the receiver functions may mainly be caused by the lower crust under the LMS area. Moreover, geological studies (e.g. Xu et al., 2008; Wang and Meng, 2009) show a strong shear deformation along the western and northern margins of the LMS belt. Rock recipe modeling (Tatham et al., 2008) demonstrated that mica and amphibole as two strongly anisotropic crustal minerals can generate a large seismic anisotropy associated with lattice preferred orientation under strong shear through deformation (Crampin, 1984; Silver, 1996). This implies that the principal cause of the observed significant crustal anisotropy may be a preferential orientation of mica and amphibole induced by the lower crustal deformation in the study area. We note that stations YZP and GAZ located on the LMS belt, and between the Xianshuihe, Xiaojiang and LMS strike-slip faults, respectively, each exhibits a significant splitting time. The fast polarization direction under station YZP is parallel to the LMS fault (Fig. 7), while under station GZA, it is perpendicular to the Xianshuihe fault but at a low angle to the LMS belt. However, the fast directions under these two stations are inconsistent with that of other neighboring stations. Geophysical studies (e.g. Tapponnier et al., 2001; Shi et al., 2013) showed that this region is dominated by extensive deformation and complex strike slip stress. The regional deformation includes 10–11 mm/yr left slip across the Xianshuihe fault, 7 mm/yr left slip across the Xiaojiang fault zone, and 4–6 mm/yr right slip across the LMS fault (Shen et al., 2005). Thus, the significant crustal anisotropy may relate to the combined effect of extensive stress and accumulated simple shear under these two stations. 4.3. Comparison of crustal anisotropy with GPS velocity, SKS splitting and APM Many studies (e.g., Kreemer et al., 2003; Meltzer et al., 2007; Wang et al., 2008; Burchfiel et al., 2008; Sun et al., 2012; Chen et al., 2013) have investigated the deformation in eastern Tibet and its surroundings using the surface geologic features, GPS velocity, APM, and SKS splitting analysis. The observation of GPS velocity, Moho Ps splitting, SKS splitting, and APM are generally serve as proxies for surface motion velocity, crustal deformation, mantle deformation and asthenospheric flow (Wolfe and Solomon, 1998), respectively. In order to study the mechanical deformation under the LMS area, we compare the estimated crustal anisotropy with these indicators of deformation (Fig. 7a and b). As shown in Fig. 7a, the measured fast polarization direction of crustal anisotropy is dominantly NW–SE around the LMS area. GPS measurements (Fig. 7a, Shen et al., 2005; Burchfiel et al., 2008) showed that in the eastern Tibet, the vectors orient northeast with the displacement rate of 15–20 mm/yr, while little of the northeastward convergence occurs across the LMS (<3 mm/yr) (Burchfiel et al., 2008). The comparison with the GPS measurements indicates that the deformation between the upper crust and lower crust is decoupled, and the direction of deformation in lower crust is inconsistent with that in upper crust under the LMS area. Fig. 7b shows that most of stations with the fast polarization directions of crustal anisotropy around the LMS area are approximately parallel or at a low angle to those measured from SKS data (Wang et al., 2008; Chang et al., 2008) and APM (Kreemer et al., 2003). The good agreements of the three directions (Moho Ps splitting, SKS splitting, and APM) indicate the consistent deformation between the lower crust and upper mantle. The striking parallelism between structural crustal features, the surface strain field and the fast shear-wave polarization directions beneath southeastern Tibet (Holt, 2000; Sol et al., 2007) suggested that the vertically coherent deformation can reflect either crust–mantle mechanical

19

coupling, or that the crust and mantle are subjected to similar velocity boundary conditions. Moreover, recent reports of seismic tomography (Huang and Zhao, 2006; Lei and Zhao, 2009; Zhang et al., 2009), magnetotelluric imaging (Zhao et al., 2008), and density structure (Zhang et al., 2014) showed a prominent low seismic velocity, low electrical resistivity and low density layer in the middle and lower crust beneath the LMS belt zone. If a low wave velocity or low density layer exists in middle-lower crust, the lithospheric mantle and crust can deform differently as no substantial stress transfer can occur between them (Clark and Royden, 2000; Beaumont et al., 2001). Thus, the consistent deformation between crust and mantle derived from Ps and SKS splitting data in this study may result from similar velocity boundary conditions, rather than the mechanical coupling. It suggests that the tendency of deformation in the lower crust is consistent with that of upper mantle. This interpretation is supported by a dynamic model with deformable boundaries (Bendick and Flesch, 2007) and a 3D viscous flow model (Yang and Liu, 2009) in Tibetan Plateau. Furthermore, a station-by-station comparison of the fast direction and splitting time between our measurements and the SKS results (Fig. 7b and Table 1S) show that the average splitting time (0.57 s) of our measurements is close to a half of the average splitting time (1.11 s) measured from the SKS splitting (Chang et al., 2008). Nagaya et al. (2008) studied crustal anisotropy in Japan and suggested that crustal anisotropy splitting time ranging from 0.2 to 0.7 s should be considered in interpreting SKS splitting data. According to the laboratory measurements of olivine lattice in upper mantle with an anisotropy degree of 0.04 (McNamara et al., 1994), Chang et al. (2008) deduced that the anisotropic thickness of mantle lithosphere is about 120 km based on the average splitting time 1.11 s in southeastern Tibet. However, recent studies with ScS reverberation data (Niu, 2011; Hu et al., 2011) found that the lithosphere of the LMS region is very thin (80– 120 km). Given the observed average crustal thickness is 50 km (Fig. 7a; Table 1S), the mantle lithosphere under the LMS area is only 30–70 km. It can cause seismic anisotropy with the splitting time of 0.26–0.61 s, which is much smaller than 1.11 s. Thus, the observed significant crustal anisotropy (0.57 s) in lower crust may be an important source for the shear wave splitting time measured by SKS data. In the Sichuan Basin, a low splitting times has been observed at three stations with the fast directions roughly aligned with the APM (Fig. 7a and b). The Sichuan Basin is a part of the Yangtze craton, which is a stable continental plate without obvious ruptures and deformation since it was formed in the Proterozoic (e.g., Yin and Harrison, 2000). P-wave tomography (Li et al., 2006) revealed a seismically fast structure beneath the Sichuan Basin extending to 250 km depth, indicating that the basin is underlain by a deep cold root. Thus, the low splitting time of crustal anisotropy and the consistent directions (Moho Ps splitting and APM) suggest that the Sichuan block is a stable basin with a vertically coherent deformation between the crust and mantle. 4.4. Crustal anisotropy and deformation Comparing with crustal anisotropy between the west and east of the LMS belt, we find that crustal anisotropy is quite different across the LMS fault zone (Fig. 7). Under the west side of the LMS belt, crustal anisotropy is significant with the fast direction of NW–SE, whereas less or even no crustal anisotropy has been observed under the Sichuan Basin. Moreover, the observed significant seismic anisotropy not only exists along the major fault, but also occurs within the blocks in the west of LMS area. In addition, relatively high Vp/Vs ratios are located within the thickened crust. These seismic observations imply that the deformation may be distributed within the lower crust beneath the LMS area, probably in

20

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

Fig. 8. Maps show the Moho topography (a) and the lateral variations of Vp/Vs ratio (b) with crustal anisotropy. In particular, note that a thicker Moho depth (>50 km) and higher Vp/Vs ratio (1.74–1.86) with significant crustal anisotropy in the west of the LMS compared with that of Sichuan Basin (40 km in Moho depth and 1.68 in Vp/Vs ratio) with a weak significant anisotropy.

the form of lower crustal flow as suggested by previous studies (Clark and Royden, 2000; Royden et al., 1997). This model is also in accordance with other geophysical observations in this region, e.g., magnetotelluric imaging (Zhao et al., 2008), seismic tomography (Lei and Zhao, 2009; Huang and Zhao, 2006; Zhang et al., 2009), shear wave splitting (Wang et al., 2008; Chang et al., 2008; Chen et al., 2013) and density structure (Zhang et al., 2014).

Geophysical studies (Clark et al., 2005; Chang et al., 2008; Shi et al., 2013) showed that this lower crustal flow rotates to south and/or north due to the continuous extrusion of eastern Tibet and the resistance of the rigid Sichuan Basin. However, the fast polarization directions of Moho Ps splitting show no evidence for these rotations in our results. Recent body wave tomography (Lei and Zhao, 2009) and a reflection/refraction seismic profile (Jia

Fig. 9. (a) Elevation along the profile AA’ in Fig. 1b. (b) Cartoon of crustal structure and deformation style near the LMS belt as modified from Jia et al. (2014) along the profile AA’. Red dash lines represent main faults. White arrow and black arrows illustrate the direction of the lower crustal flow and the resistance of Sichuan Basin, respectively. Black dash arrows indicate the flow of material upward along the main fault obstructed by the Sichuan Basin. Red Squares on the surface represent seismic stations in the profile of AA’, while the black bars represent the splitting time and fast polarization direction. Abbreviations are: F1: Guanxian–Jiangyou fault; F2: Yingxiu–Beichuan fault; F3: Wenchuan–Maowen. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

et al., 2014) across the LMS belt studies showed that a distinct low-velocity layer goes upward from lower crust to upper crust beneath the LMS belt, indicating that the lower crustal flow may thrust upwelling in this region. In addition, the effective elastic thickness (Te) (Fielding and McKenzie, 2012; Chen et al., 2014) showed that Te is generally lower in the LMS belt with a thickness of only 5–15 km compared to 30–40 km in the Sichuan Basin, suggesting that the lithosphere is weak beneath the LMS area. Combined with the observed thickened crust with high Vp/Vs ratio in this region, we speculate that the lower crust flow may accumulate in the lower crust and partly extrude into the upper crust under the LMS belt (Fig. 9). Geological studies (e.g. Xu et al., 2008; Liu-zeng et al., 2009) suggest that some steeply dipping faults (e.g. Wenchuan–Maowen, Yingxiu–Beichuan and Guanxian–Jiangyou faults) stretch down to the lower crust. These observations imply that this upwelling flow along the eastern margin of the orogenic belt may be associated with these steeply dipping thrusts, resulting in the surface uplift of the LMS belt. 5. Conclusions The crustal structure and seismic anisotropy beneath the LMS and its surroundings have been investigated through the analysis of receiver function data. We find significant crustal anisotropy with a relatively high Vp/Vs ratio (1.74–1.86) and thickened crust (50–67 km) under the LMS area. Both the fast polarization directions and splitting times of Moho Ps splitting are compared to those measured from SKS splitting, suggesting that the lower crustal anisotropy is an importance source of SKS splitting and the crust–mantle consistent deformation in the LMS area, and strongly deformation is distributed in the lower crust. All those observations are consistent with a scenario of lower crustal extrusion beneath the LMS area. Combined with the recent seismic reflection/refraction profile and geology studies, we suggest that the lower crustal flow may extrude upward into the upper crust along steeply dipping thrusts under the LMS belt, due to the continuous extrusion of Tibetan materials and the resistance of the rigid Sichuan Basin, causing the surface uplift of the LMS area. Acknowledgments We thank the Data Management Center of the China Earth quake Administration for providing the waveform data for this study. We thank Fenglin Niu for providing the estimated crustal anisotropy method and Bob King and B. Clark Burchfiel for providing the GPS velocity data in this study. We are very grateful to two anonymous reviewers and the editor, Dr. Vernon F. Cormier, for their critical review and constructive comments, which significantly improved the quality of this paper. We also thank Prof. David Nobes for improving the English expression. This study is supported by National Natural Science Foundation of China (Grant Nos. 41404042, 41274099 and 41404061) and Postdoctoral Science Foundation of China (Grant No. 2014M552147). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.pepi.2015.04.005. References Ammon, C.J., 1991. The isolation of receiver effect from teleseismic p waveforms. Bull. Seismol. Soc. Am. 81, 2504–2510.

21

Beaumont, C., Jamieson, R.A., Nguyen, M.H., Lee, B., 2001. Himalayan tectonics explained by extrusion of a low-viscosity crustal channel coupled to focus surface denudation. Nature 414 (6865), 738–742. Bendick, R., Flesch, L., 2007. Reconciling lithospheric deformation and lower crustal flow beneath central Tibet. Geology 35 (10), 895–898. 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(7), p. 5. Chang, L., Wang, C., Ding, Z., 2008. Seismic anisotropy of upper mantle in Sichuan and adjacent regions. Sci. China Ser. D Earth Sci. 51 (12), 1683–1693. Chen, Y., Badal, J., Hu, J.F., 2010. Love and Rayleigh wave tomography of the QinghaiTibet Plateau and surrounding areas. Pure Appl. Geophys. 167, 1171–1203. Chen, Y., Zhang, Z., Sun, C., Badal, J., 2013. Crustal anisotropy from Moho converted Ps wave splitting analysis and geodynamic implications beneath the eastern margin of Tibet and surrounding regions. Gondwana Res. 24 (3), 946–957. Chen, B., Liu, J., Kaban, M.K., Sun, Y., Chen, C., Du, J., 2014. Elastic thickness, mechanical anisotropy and deformation of the southeastern Tibetan Plateau. Tectonophysics 637, 45–56. Christensen, N.I., 1996. Poisson’s ratio and crustal seismology. J. Geophys. Res. 101, 3139–3156. Clark, M.K., Royden, L.H., 2000. Topographic ooze: building the eastern margin of Tibet by lower crustal flow. Geology 28 (8), 703–706. Clark, M.K., Bush, J.W., Royden, L.H., 2005. Dynamic topography produced by lower crustal flow against rheological strength heterogeneities bordering the Tibetan Plateau. Geophys. J. Int. 162 (2), 575–590. Clayton, R.W., Wiggins, R.A., 1976. Source shape estimation and deconvolution of teleseismic bodywaves. Geophys. J. Int. 47 (1), 151–177. Crampin, S., 1984. Evaluation of anisotropy by shear wave splitting. Applied Seismic Anisotropy: Theory, Background, and Field Studies, Geophysics Reprint series, 20, 23–33. Crampin, S., Peacock, S., 2008. A review of the current understanding of seismic shear-wave splitting in the Earth’s crust and common fallacies in interpretation. Wave Motion 45 (6), 675–722. Fielding, E.J., McKenzie, D., 2012. Lithospheric flexure in the Sichuan Basin and Longmen Shan at the eastern edge of Tibet. Geophys. Res. Lett. 39, L09311. http://dx.doi.org/10.1029/2012GL051680. Gao, S., Zhang, B.R., Jin, Z.M., Kern, H., Luo, T.C., Zhao, Z.D., 1998. How mafic is the lower continental crust? Earth Planet. Sci. Lett. 161 (1), 101–117. Holt, W.E., 2000. Correlated crust and mantle strain fields in Tibet. Geology 28 (1), 67–70. Houseman, G., England, P., 1993. Crustal thickening versus lateral expulsion in the Indian Asian continental collision. J. Geophys. Res. 98 (B7), 12233–12249. Hu, J., Xu, X., Yang, H., Wen, L., Li, G., 2011. S receiver function analysis of the crustal and lithospheric structures beneath eastern Tibet. Earth Planet. Sci. Lett. 306 (1), 77–85. Huang, J., Zhao, D., 2006. High-resolution mantle tomography of China and surrounding regions. J. Geophys. Res. 111 (B9), B09305. http://dx.doi.org/ 10.1029/2005JB004066. Jia, S.X., Liu, B.J., Xu, Z.F., et al., 2014. The crustal structures of the central Longmenshan along and its margins as related to the seismotectonics of the 2008 Wenchuan Earthquake. Sci. China Earth Sci. 57, 777–790. http:// dx.doi.org/10.1007/s11430-013-4744-9. Kawakatsu, H., Niu, F., 1994. Seismic evidence for a 920-km discontinuity in the mantle. Nature 371, 301–305. Kennett, B.L.N., Engdahl, E.R., 1991. Travel times for global earthquake location and phase identification. Geophys. J. Int. 105, 429–465. Kreemer, C., Holt, W.E., Haines, A.J., 2003. An integrated global model of present-day plate motions and plate boundary deformation. Geophys. J. Int. 154, 8–34. Lei, J., Zhao, D., 2009. Structural heterogeneity of the Longmenshan fault zone and the mechanism of the 2008 Wenchuan earthquake (Ms 8.0). Geochem. Geophys. Geosyst. 10, Q10010. http://dx.doi.org/10.1029/2009GC002590. Lev, E., Long, M.D., van der Hilst, R.D., 2006. Seismic anisotropy in Eastern Tibet from shear wave splitting reveals changes in lithospheric deformation. Earth Planet. Sci. Lett. 251 (3), 293–304. Levin, V., Park, J., 1997. P-SH conversions in a flat-layered medium with anisotropy of arbitrary orientation. Geophys. J. Int. 131, 253–266. Li, C., van der Hilst, R.D., Toksoz, N.M., 2006. Constraining spatial variations in Pwave velocity in the upper mantle beneath SE Asia. Phys. Earth Planet. Inter. 154, 180–195. Liu, H., Niu, F., 2012. Estimating crustal seismic anisotropy with a joint analysis of radial and transverse receiver function data. Geophys. J. Int. 188, 144–164. Liu-Zeng, J., Zhang, Z., Wen, L., Tapponnier, P., Sun, J., Xing, X., Van der Woerd, J., 2009. Co-seismic ruptures of the 12 May 2008, M s 8.0 Wenchuan earthquake, Sichuan: East–west crustal shortening on oblique, parallel thrusts along the eastern edge of Tibet. Earth Planet. Sci. Lett. 286 (3), 355–370. Mainprice, D., Nicolas, A., 1989. Development of shape and lattice preferred orientations: application to the seismic anisotropy of the lower crust. J. Struct. Geol. 11, 175–189. Makovsky, Y., Klemperer, S.L., Ratschbacher, L., Brown, L.D., Li, M., Zhao, W.J., Meng, F.L., 1996. INDEPTH wide-angle reflection observation of P-wave-to-S-wave conversion from crustal bright spots in Tibet. Science 274, 1690–1691. McNamara, D.E., Owens, T.J., 1993. Azimuthal shear wave velocity anisotropy in the Basin and Range Province using Moho Ps converted phases. J. Geophys. Res. 98 (12), 003–12017.

22

Y. Sun et al. / Physics of the Earth and Planetary Interiors 244 (2015) 11–22

McNamara, D., Owens, T., Silver, P.G., et al., 1994. Shear-wave anisotropy beneath the Tibetan Plateau. J. Geophys. Res. 99, 13655–13665. Meltzer, A.S., Bürgmann, R., van der Hilst, R.D., King, R., Chen, Z., Koons, P.O., Lev, E., Liu, Y.P., Zeitle, P.K., Zhang, X., Zhang, J., Zurek, B., 2007. Geodynamics of the southeastern Tibetan Plateau from seismic anisotropy and geodesy. Geology 35 (6), 563–566. Molnar, P., Tapponnier, P., 1975. Cenozoic tectonics of Asia: effects of a continental collision. Science 189 (4201), 419–426. Nagaya, M., Oda, H., Akazawa, H., Ishise, M., 2008. Receiver functions of seismic waves in layered anisotropic media: application to the estimate of seismic anisotropy. Bull. Seismol. Soc. Am. 98 (6), 2990–3006. Ni, J.F., Guzman-Speziale, M., Bevis, M., Holt, W.E., Wallace, T.C., Seager, W.R., 1989. Accretionary tectonics of Burma and the three-dimensional geometry of the Burma subduction zone. Geology 17 (1), 68–71. Nicolas, A., Christensen, N.I., 1987. Formation of anisotropy in upper mantle peridotites-a review. In: Fuchs, K., Froideveaux, C. (Eds.), Composition Structure and Dynamics of the Lithosphere–Asthenosphere System, Geodynamics Series, vol. 16. American Geophysical Union, Washington, D.C., pp. 111–123. Niu, F., 2011. Mapping Lithosphere Thickness Beneath China with ScS Reverberation Data: What Controls the Intraplate Seismicity in Mainland China, Abstract S23E-02 presented at 2011 Fall Meeting, AGU, San Francisco, Calif., 5–9 Dec. Niu, F., James, D.E., 2002. Fine structure of the lowermost crust beneath the Kaapvaal craton and its implications for crustal formation and evolution. Earth Plant. Sci. Lett. 200, 121–130. Niu, F., Kawakatsu, H., 1998. Determination of the absolute depths of the mantle transition zone discontinuities beneath China: effect of stagnant slabs on mantle transition zone discontinuities. Earth Plant. Sci. Lett. 50, 965–975. Niu, F., Li, J., 2011. Component azimuths of the CEArray stations estimated from P wave particle motion. Earthquake Sci. 24, 3–13. Niu, F., Baldwin, T., Pavlis, G., Vernon, F., Rendon, H., Bezada, M., Levander, A., 2007. Receiver function study of the crustal structure of the Southeastern Caribbean plate boundary and Venezuela. J. Geophys. Res. 112, B11308. Owens, T.J., Zandt, G., 1997. Implications of crustal property variations for models of Tibetan plateau evolution. Nature 387 (6628), 37–43. Peng, X., Humphreys, E.D., 1997. Moho dip and crustal anisotropy in northwestern Nevada from teleseismic receiver functions. Bull. Seismol. Soc. Am. 87 (3), 745– 754. Royden, L.H., Burchfiel, B.C., King, R.W., Wang, E., Chen, Z., Shen, F., Liu, Y., 1997. Surface deformation and lower crustal flow in eastern Tibet. Science 276 (5313), 788–790. Shen, Z.K., Lü, J., Wang, M., Bürgmann, R., 2005. Contemporary crustal deformation around the southeast borderland of the Tibetan Plateau. J. Geophys. Res. 110 (B11). Shi, Y.T., Gao, Y., Zhang, Y.J., et al., 2013. Shear wave splitting in the crust in Eastern Songpan–Garze block Sichuan–Yunnan block and Western Sichuan Basin. Chin. J. Geophys. (in Chin.) 56 (Z), 481–494. Shiomi, K., Park, J., 2008. Structural features of the subducting slab beneath the Kii Peninsula, central Japan: seismic evidence of slab segmentation, dehydration, and anisotropy. J. Geophys. Res. 113 (B10). Silver, P.G., 1996. Seismic anisotropy beneath the continents: probing the depths of geology. Annu. Rev. Earth Planet. Sci. 24, 385–432. Sol, S., Meltzer, A., Bürgmann, R., Van der Hilst, R.D., King, R., Chen, Z., Zurek, B., 2007. Geodynamics of the southeastern Tibetan Plateau from seismic anisotropy and geodesy. Geology 35 (6), 563–566. Sun, Y., Niu, F., Liu, H., Chen, Y., Liu, J., 2012. Crustal structure and deformation of the SE Tibetan plateau revealed by receiver function data. Earth Planet. Sci. Lett. 349, 186–197.

Tapponnier, P., Peltzer, G., Le Dain, A.Y., Armijo, R., Cobbold, P., 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticize. Geology 10, 611–616. Tapponnier, P., Zhiqin, X., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G., Jingsui, Y., 2001. Oblique stepwise rise and growth of the Tibet Plateau. science 294 (5547), 1671–1677. Tatham, D.J., Lloyd, G.E., Butler, R.W.H., Casey, M., 2008. Amphibole and lower crustal seismic properties. Earth Planet. Sci. Lett. 267, 118–128. Wang, E., Meng, Q., 2009. Mesozoic and Cenozoic tectonic evolution of the Longmenshan fault belt. Sci. China, Ser. D Earth Sci. 52 (5), 579–592. Wang, C.Y., Flesch, L.M., Silver, P.G., Chang, L.J., Chan, W.W., 2008. Evidence for mechanically coupled lithosphere in central Asia and resulting implications. Geology 36 (5), 363–366. Watson, M., Hayward, A.B., Parkinson, D.N., Zhang, Z.M., 1987. Plate tectonic history, basin development and petroleum source rock deposition onshore China. Mar. Pet. Geol. 4, 205–225. Wei, W., Zhao, D., Xu, J., 2013. P-wave anisotropic tomography in Southeast Tibet: new insight into the lower crustal flow and seismtectonics. Phys. Earth Planet. Inter. 222, 47–57. Wolfe, C.J., Solomon, S.C., 1998. Shear-wave splitting and implications for mantle flow beneath the MELT region of the East Pacific Rise. Science 280 (5367), 1230– 1232. Xu, Y.G., Menzies, M.A., Thirlwall, M.F., Xie, G.H., 2001. Exotic lithosphere mantle beneath the western Yangtze craton: petrogenetic links to Tibet using highly magnesian ultrapotassic rocks. Geology 29 (9), 863–866. Xu, L., Rondenay, S., van der Hilst, R.D., 2007. Structure of the crust beneath the southeastern Tibetan plateau from teleseismic receiver functions. Phys. Earth Planet. Inter. 165, 176–193. Xu, Z., Ji, S., Li, H., Hou, L., Fu, X., Cai, Z., 2008. Uplift of the Longmen Shan range and the Wenchuan earthquake. Episodes 31 (3), 291. Yang, Y., Liu, M., 2009. Crustal thickening and lateral extrusion during the IndoAsian collision: a 3D viscous flow model. Tectonophysics 465 (1), 128–135. Yao, H., Van der Hilst, R.D., Montagner, J.P., 2010. Heterogeneity and anisotropy of the lithosphere of SE Tibet from surface wave array tomography. J. Geophys. Res. 115, B12307. Yin, A., Harrison, T.M., 2000. Geologic evolution of the Himalayan-Tibetan orogen. Annu. Rev. Earth Planet. Sci. 28, 211–280. Yong, L., Allen, P.A., Densmore, A.L., Qiang, X., 2003. Evolution of the Longmen Shan foreland basin (western Sichuan, China) during the Late Triassic Indosinian orogeny. Basin Res. 15 (1), 117–138. Zhang, H.F., Zhang, B.R., Harris, N., Zhang, L., Chen, Y.L., Chen, N.S., Zhao, Z.D., 2006. U-Pb zircon SHRIMP ages, geochemical and Sr–Nd–Pb isotopic compositions of intrusive rocks from the Longshan–Tianshui area in the southeast corner of the Qilian orogenic belt, China: constraints on petrogenesis and tectonic affinity. J. Asian Earth Sci. 27 (6), 751–764. Zhang, Z., Wang, Y., Chen, Y., Houseman, G.A., Tian, X., Wang, E., Teng, J., 2009. Crustal structure across Longmenshan fault belt from passive source seismic profiling. Geophys. Res. Lett. 36 (17). Zhang, Y., Teng, J., Wang, Q., et al., 2014. Density structure and isostatic state of the crust in the Longmenshan and adjacent areas. Tectonophysics 619, 51–57. Zhao, G., Chen, X., Wang, L., Wang, J., Tang, J., Wan, Z., Xiao, Q., 2008. Evidence of crustal ‘channel flow’ in the eastern margin of Tibetan Plateau from MT measurements. Chin. Sci. Bull. 53 (12), 1887–1893. Zheng, X.F., Yao, Z.X., Liang, J.H., Zheng, J., 2010. The role played and opportunities provided by IGP DMC of China National Seismic Network in Wenchuan earthquake disaster relief and researches. Bull. Seismol. Soc. Am. 100 (5B), 2866–2872. Zhu, L., Kanamori, H., 2000. Moho depth variation in southern California from teleseismic receiver functions. J. Geophys. Res. 105, 2969–2980.