Seismic tomographic imaging of the crust and ... - Wiley Online Library

9 downloads 14 Views 5MB Size Report
Center) seismic stations located in western Tien Shan and north of Lake Issyk‐Kul ..... 1998]. I call the earthquakes after 1964 the “recent earth- quakes. ..... Bielinski, R., S. Park, A. Rybin, V. Batalev, S. Jun, and C. Sears (2003),. Lithospheric ...

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 116, B09305, doi:10.1029/2010JB008000, 2011

Seismic tomographic imaging of the crust and upper mantle under the central and western Tien Shan orogenic belt Jianshe Lei1 Received 16 September 2010; revised 4 June 2011; accepted 24 June 2011; published 16 September 2011.

[1] Detailed 3‐D tomographic images of P and S wave velocity (Vp, Vs) and Poisson’s

ratio (s) under the central and western Tien Shan orogenic belt are determined by using a large number of high‐quality P and S wave arrival times from local earthquakes. The results show that under the Tien Shan orogenic belt high‐Vp, high‐Vs, and low‐s anomalies are revealed in the upper and middle crust, possibly indicating the existence of the Paleozoic crystalline basement rocks, while low‐Vp, low‐Vs, and high‐s anomalies appear in the lower crust and upper mantle, perhaps suggesting that the hot and wet material is upwelling under the Tien Shan orogenic belt from the mantle. Some high‐Vp, high‐Vs, and low‐s anomalies are tilted toward the Tien Shan along with the seismicity. These are found in the collision zones between the Tien Shan and the Tarim basin in the south and the Kazakh shield in the north and suggest the underthrusting of the Tarim and Kazakh lithosphere beneath the Tien Shan. Meanwhile, some low‐Vp, low‐Vs, and high‐s anomalies are imaged in other parts of these collision zones, perhaps indicating the intrusion of the hot and wet material into the crust from the upper mantle. These results indicate that both the upwelling of the hot and wet material and the underthrusting of the Tarim and Kazakh lithosphere may have played an important role in the mountain building. Under the Tarim and Fergana basins, low‐Vp, low‐Vs, and high‐s anomalies are revealed in the upper crust, while high‐Vp, high‐Vs, and low‐s anomalies are visible in the lower crust and upper mantle. These may reflect the existence of less compacted sedimentary material in the shallow crust and more highly compacted craton‐like structures in the deeper crust and upper mantle under the basins. The Talas‐Fergana fault shows an obvious tectonic boundary between central and western Tien Shan. The central Tien Shan displays high‐Vp, high‐Vs, and low‐s anomalies in the upper and middle crust, while western Tien Shan exhibits low‐Vp, low‐Vs, and high‐s anomalies. However, the pattern of seismic structure between central and western Tien Shan reverses in the lower crust. Such a correlation may extend down to the upper mantle, suggesting that the Talas‐Fergana fault may be a lithospheric‐scale boundary. Additionally, a columnar low‐Vp and low‐Vs anomaly is clearly observed around the turning point of the Talas‐Fergana fault from the NWN to NWW trending orientations and may indicate that the fault provides a channel for the hot and wet material upwelling from the mantle to the surface. Citation: Lei, J. (2011), Seismic tomographic imaging of the crust and upper mantle under the central and western Tien Shan orogenic belt, J. Geophys. Res., 116, B09305, doi:10.1029/2010JB008000.

1. Introduction [2] The Tien Shan mountains, located in central Asia, are about 2000 km away from the India‐Asia collision front. The range extends roughly 2500 km in the east‐west direction with a maximum width of about 400 km at its west end. This orogenic belt includes several parallel ranges and 1 Seismological Laboratory, Institute of Crustal Dynamics, China Earthquake Administration, Beijing, China.

Copyright 2011 by the American Geophysical Union. 0148‐0227/11/2010JB008000

intermontane basins and is surrounded by several stable blocks, such as the Kazakh shield to the north and the Tarim basin to the south (Figure 1). Moreover, this mountain range is one of the youngest, the highest, and the most active orogenic belts in the world and is regarded as an ideal place to study the mechanism of the active intracontinental mountain building. [3] There are many active faults with small and large earthquakes in and around the Tien Shan (Figure 1), possibly indicating that the tectonic activity resumed in the Oligocene and has continued to the present‐day [e.g., Sobel and Dumitru, 1997; Chen et al., 1999; Molnar and Ghose,

B09305

1 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 1. Topographic map of central and western Tien Shan showing the major regional tectonic features. Thin lines denote major active faults, while the thick line denotes the Talas‐Fergana fault. These traces were scanned from Ghose et al. [1998]. Dots show the earthquakes that occurred between 1964 and 2004 [Engdahl et al., 1998]. Stars represent the epicenters of large historic earthquakes (M ≥ 7.0) in the region since the late nineteenth century, compiled from Kondorskaya and Shebalin [1982]. White squares represent the study area. The insets show the geographic location and legend of the map. WTS, western Tien Shan; CTS, central Tien Shan; ETS, eastern Tien Shan. 2000] presumably as a consequence of the India‐Asia collision, though the whole area appears to have been stable throughout the Mesozoic with little topographic relief in the late Cretaceous. The orogenic belt is often divided into three portions along its length: western, central and eastern Tien Shan (Figure 1). The central and west portions of the Tien Shan are separated by the Talas‐Fergana fault, while eastern Tien Shan is located east of Lake Issyk‐Kul or 80°E, within China [e.g., Sobel and Arnaud, 2000; Lei and Zhao, 2007]. In order to better understand the mechanism of the Tien Shan mountain building, some researchers have investigated the deep structure and dynamics of eastern Tien Shan and obtained intriguing results [e.g., Poupinet et al., 2002; Zhao et al., 2003]. With the deployment of the temporary seismic network GHENGIS (The Tien Shan Continental Dynamic project) run by Roecker [2001] in central Tien Shan, many seismological results have been inferred by using receiver function analyses [e.g., Vinnik et al., 2004; Tian et al., 2010], SKS splitting observations [e.g., Li and Chen, 2006; Jiang

et al., 2010], and seismic tomography [e.g., Lei and Zhao, 2007; Z. Li et al., 2009; Omuralieva et al., 2009]. [4] Lei and Zhao [2007] used the teleseismic data collected from the digital seismograms recorded by several seismic networks including the GHENGIS, KN (Kyrgyzstan Seismic Telemetry Network), G (Geoscope) and KZ (Kazakhstan Network) networks to infer the deep mantle structure under central Tien Shan. Due to the nearly vertical raypaths to the seismic networks from the teleseismic events, it is difficult to resolve the shallow crustal structure under the region [Lei and Zhao, 2005; Lei et al., 2009a, 2009b]. Recently, Omuralieva et al. [2009] presented a crustal model developed using the arrival time data from the local earthquakes recorded by the seismic networks as used by Lei and Zhao [2007], but their work only showed the seismic structure of P and S wave velocity (Vp, Vs) under central Tien Shan. To better understand the geodynamic process of the Tien Shan mountain, I derive a new velocity model of the crust including Poisson’s ratio (s).

2 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 2. Distribution of local earthquakes (circles) and seismic stations (other symbols) used in present study. Triangle, GHENGIS; inverted triangle, KN; hexagon, KZ; square, G; diamond, ISC (International Seismic Stations). The arrival times from the ISC data sets were reprocessed by Engdahl et al. [1998]. The curves denote the outlines of the Tarim and Fergana basins, the Talas‐Fergana fault, Lake Issyk‐Kul, and the boundary between the Kazakh shield and the Tien Shan. [5] In addition to collecting the arrival time data by hand‐ picking from high‐quality seismograms of local earthquakes recorded by several seismic networks as mentioned above, I added the arrival times recorded by seismic stations west of the Talas‐Fergana fault to simultaneously unravel the detailed seismic structure of Vp, Vs and s under central and western Tien Shan orogenic belt. Because Poisson’s ratio is more sensitive to compositional variations and content of fluids and melts than Vp and Vs alone, it is a very useful parameter in studying physical and petrologic properties of the crust and upper mantle. Poisson’s ratio can be determined by following the relation (Vp/Vs)2 = 2(1 − s)/(1 − 2s) [Zhao et al., 1996]. The present results shed new insights into the geodynamic process of this enigmatic orogenic belt and seismotectonics in the region.

2. Data and Method [6] In the present study I collected the arrival time data from 1062 local earthquakes recorded by 73 seismic stations in central and western Tien Shan from October 1997 to August 2000. These networks include 10 KN, 1 KZ, 1 G, and 28 GHENGIS digital broadband stations spanned over central Tien Shan, as well as 33 ISC (International Seismic Center) seismic stations located in western Tien Shan and north of Lake Issyk‐Kul (Figure 2). The arrival time data from these 33 ISC stations were directly extracted from the bulletins that were reprocessed by Engdahl et al. [1998].

Compared to previous studies [e.g., Omuralieva et al., 2009], the present work has a wider and denser distribution of seismic stations in the region. These stations generally have an even distribution and cover the entire central and western Tien Shan (Figure 2). The stations of the GHENGIS network were equipped with three‐component broadband sensors, CMG‐3ESP or STS‐2, operating with a sampling rate of 40 Hz during October 1997 and August 2000. The stations of the remaining networks were operated for a few years to more than ten years. Except for several Chinese stations located on sediments, stations in Kyrgyzstan and Kazakhstan were installed on the crystalline or metamorphic basement rocks. For more details of these networks, see Roecker [2001] and Lei and Zhao [2007]. [7] The distribution of local earthquakes is uneven. Most of them are concentrated on the boundaries between the Tien Shan and Tarim basin in central Tien Shan and Pamir plateau in western Tien Shan, with some around the Fergana basin (Figure 1). In the data set each event has 8 to 100 recordings. Some travel time data were winnowed out because their residuals are larger than 6.0 s. In total, I collected 21,711 P wave arrival times and 12,748 S wave arrival times, over 70% of which were hand‐picked from high‐quality seismograms recorded by the GHENGIS, G, KZ and KN networks. The accuracy for the P and S wave arrival times is estimated to be 0.1 and 0.2 s, respectively. Figure 3 shows the observed travel times versus the epicentral distance and the travel time residuals before and after

3 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 3. (a and e) Observed (black dots) and calculated (gray dots) travel time curves (in s) versus epicentral distance (in km). (b and f) Travel time residuals versus epicentral distance (in km) before tomographic inversion. (c and g) Travel time residuals versus epicentral distance (in km) after tomographic inversion. (d and h) Histograms of travel time residuals before (dashed lines) and after (solid lines) the inversion. the inversions. It can be seen that most of the travel‐time data for both P waves and S waves are concentrated in a range of epicentral distance less than 500 km, with rather more P wave than S wave data beyond 500 km. [8] I applied the tomographic technique of Zhao et al. [1992] to the P and S wave arrival times to determine a detailed 3‐D crustal and upper mantle structure under central and western Tien Shan. This method calculates the travel times and raypaths accurately and efficiently, and treats a model containing complex seismic discontinuities. For details, see Zhao et al. [1992]. In the present study I used 1‐D Vp and Vs velocity models (Figure 4) containing an undulating Moho discontinuity. The 1‐D velocity models

in the crust are inferred from previous studies [e.g., Vinnik et al., 2004; Bassin et al., 2000], but those in the mantle are adopted from the iasp91 velocity models [Kennett and Engdahl, 1991]. The Moho discontinuity under central Tien Shan directly is adopted from the model of Vinnik et al. [2004] that was derived from receiver function analyses, while outside of central Tien Shan the Moho geometry is taken from the model CRUST2.0 [Bassin et al., 2000] (http:// igpweb.ucsd.edu/~gabi/crust2.html) which is an updated version of CRUST5.1 [Mooney et al., 1998] and is specified on a 2° × 2° grid in the horizontal directions. [9] A 3‐D grid was set up in the model to specify the velocity structure. The model was parameterized with an

4 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 4. (a) Initial 1‐D velocity models used in the present study. Solid line denotes P wave velocity model, while dashed line denotes S wave velocity model. In the crust, the models under central Tien Shan are inferred from receiver function analyses [Vinnik et al., 2004], while those outside of central Tien Shan are obtained from the CRUST2.0 model [Bassin et al., 2000]. In the mantle, the models are adopted directly from the iasp91 1‐D velocity model [Kennett and Engdahl, 1991]. Vp/Vs = 1.73. (b) Depth variation of the Moho discontinuity in the studying region. Under central Tien Shan, the Moho was inferred by Vinnik et al. [2004], while outside of central Tien Shan it was adopted from CRUST2.0 model [Bassin et al., 2000]. The thick line denotes the Talas‐Fergana fault, while others denote the outlines of the Tarim and Fergana basins, and Lake Issyk‐Kul, and boundary between the Kazakh shield and the Tien Shan. optimal grid spacing of 0.5° × 0.5° in the horizontal directions and 1, 15, 30, 45, 65 and 90 km were taken as layer depths (Figures 5–8). Velocity perturbations at the grid nodes were taken as unknown parameters. Figures 5 and 6 show hit counts (number of rays passing around each grid node) for P and S waves around the Tien Shan in map view, respectively. Hit counts for P waves are much higher than those for S waves, but hit count variations for P and S waves with depth are similar. At shallow depths hit counts are very uneven and closely related to the distribution of the seismic stations (Figures 5a–5c and 6a–6c), but with increasing depth hit counts gradually become uniform (Figures 5d, 5e, 6d, and 6e). The hit counts for both P and S wave data decrease rapidly at 90 km depth (Figures 5f and 6f). I resolved 1825 Vp perturbations and 1636 Vs perturbations at the grid nodes with hit counts larger than 10. During the inversion, the Moho depth was fixed and the earthquakes were relocated by using both P and S wave arrival times. The conjugate‐ gradient algorithm LSQR [Paige and Saunders, 1982] with damping and smoothing regularizations [Zhao, 2004; Lei and Zhao, 2006] was used to resolve the large and sparse systems of observational equations. For more details of the tomographic method, see Zhao et al. [1992].

3. Resolution Analyses and Results 3.1. Resolution Analyses [10] The checkerboard resolution test is used to examine the resolution scale of seismic data and evaluate the ray

coverage in the study region. Many checkerboard resolution tests were carried out by changing the grid spacing of the model. In this paper I show only the results of the checkerboard resolution tests with two different grid intervals of 0.5° × 0.5° and 1.0° × 1.0° in the horizontal directions, respectively (Figures 7–10). Alternative negative and positive velocity anomalies of up to ±5% are assigned to the 3‐D grid nodes in the input model. Random noise with zero mean and standard deviations of 0.1 and 0.2 s is added to the synthetic travel times of P and S waves, respectively, to account for the data errors which are usually present in a real data set. Output models are obtained by using the same algorithm and the same numbers of seismic stations, events, and raypaths as in the real inversion. The resolution is good for the areas where the checkerboard images are reconstructed. Figures 7 and 8 show the results of the checkerboard resolution tests for P and S waves with a grid interval of 0.5° × 0.5° in the horizontal directions (hereafter called the 0.5° test). At 1 km depth the resolution for S waves is better than that for P waves (Figures 7a and 8a) in central and western Tien Shan and Tarim basin, but at 15 km depth the resolution for both P and S waves is good in western Tien Shan and small portions of central Tien Shan (Figures 7b and 8b). However, with increasing depth the resolution for P and S waves is significantly improved in the entire study region (Figures 7c–7e and 8c–8e). At 90 km depth the good resolution area is biased toward the western Tien Shan and the Tarim basin (Figures 7f and 8f). These

5 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 5. Distribution of the number of P wave raypaths passing through each grid node (hit counts) in map view. The layer depth is shown above each map. The scale for hit counts in the square root (SQRT) is shown at the bottom. Other symbols are the same as shown in Figure 2. variations are in general agreement with those of hit counts (Figures 5 and 6). [11] Figures 9 and 10 show the results of the checkerboard tests for P and S waves with a horizontal grid interval of 1.0° × 1.0°(hereafter called the 1.0° test). Although the amplitudes of velocity anomalies are not fully retrieved, the patterns of alternative positive and negative anomalies are generally resolved in the region (Figures 9 and 10), which is much better than those of the 0.5° tests (Figures 7 and 8). At 1 km depth, in addition to central and western Tien Shan and the Tarim basin, the 1.0° tests for P and S wave structures show a good pattern of resolved anomalies in the Kazakh shield and the Pamir plateau (Figures 9a and 10a) that is not illustrated in Figures 7a and 8a. At 15 km depth, the resolution for both P and S waves (Figures 9b and 10b) shows a much better pattern of velocity anomalies in the entire study region, which is much improved in the 0.5° tests (Figure 7b and 8b). At depths of 30 and 45 km, the good‐

resolution area for both P and S waves (Figures 9c, 9d, 10c, and 10d) extends from central and western Tien Shan as shown in Figures 7c, 7d, 8c, and 8d to the Tarim basin and Kazakh shield. At 65 km depth, the resolution for P wave structure is improved in the Tarim basin (Figure 9e), while for S wave structure it is improved in the Pamir plateau (Figure 10e). At 90 km depth, the pattern of velocity anomalies in the 1.0° tests for P and S waves (Figures 9f and 10f) is significantly improved in central Tien Shan and the Pamir plateau, respectively, compared to that of the 0.5° tests (Figures 7f and 8f). 3.2. Tomographic Results [12] Figure 3 compares the travel‐time residuals before and after the tomographic inversions. It can be seen that both P and S wave travel‐time residuals for my final 3‐D tomographic models are significantly reduced relative to those for the 1‐D velocity models (Figures 3d and 3h). For

6 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 6. The same as Figure 5 but for S wave raypaths.

P wave travel times, the root‐mean square (RMS) travel time residuals are reduced by 24% from 2.35 s before the inversion to 1.90 s after the inversion, while for S wave travel times, the RMS residuals are reduced by 29% from 2.43 s to 1.88 s. [13] Figure 11 displays the final tomographic images of Vp, Vs and s under central and western Tien Shan in map view. Here, I only show the s images above 65 km in map view due to the lack of S rays below this depth (Figure 6f). Figures 12 and 13 show the resulting tomographic images along four vertical cross sections. The s model is calculated from the inverted Vp and Vs models, so that the s model is relatively less reliable than the velocity model, but it shows important structural features. [14] Some common structural features have been revealed under the Tarim and Fergana basins. Obvious low‐Vp, low‐ Vs, and high‐s anomalies are inferred in the shallow crust (Figures 11a, 11b, 11g, 11h, 11m, 11n, 12, 13a, 13b, and 13d–13f), and remarkable high‐Vp, high‐Vs and low‐s

anomalies appear in the mid‐lower crust and upper mantle (Figures 11c, 11d, 11i, 11j, 11o, 11p, 12b, 12c, 12e, 12f, 13a, 13b, 13d, and 13e). These results suggest the existence of Cenozoic sediments in the shallow crust, and stable craton‐like structure at greater depths under the basins. Low‐s anomalies under the Fergana basin are confined to only about 30 km depth (Figures 11o and 13f), which may reflect the thinner lithosphere of the Fergana basin. [15] Under the central Tien Shan a high‐Vp, high‐Vs and low‐s anomaly is imaged in the upper and middle crust, while prominent low‐Vp, low‐Vs and high‐s anomalies are revealed in the lower crust and upper mantle (Figures 11– 13). These results are generally consistent with those from local and teleseismic tomography [e.g., Lei and Zhao, 2007; Omuralieva et al., 2009], indicating the existence of the Paleozoic crystalline basement rocks in the shallow crust and the upwelling of the hot and wet material from the mantle. Most of large earthquakes occurred in and around the zones with high‐Vp, high‐Vs and low‐s anomalies, but their

7 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 7. Results of the checkerboard resolution test for P wave structure in map view. Open and solid circles denote low and high velocity anomalies. The scale for the velocity perturbation (in %) is shown at the bottom. The grid spacing is 0.5° × 0.5° in the horizontal directions. The layer depth is shown above each map. Other symbols are the same as shown in Figure 2. hypocentersare underlain by low‐Vp, low‐Vs and high‐s anomalies (Figures 11–13). [16] There exists an obvious structural contrast between the eastern and western portions of Lake Issyk‐Kul (Figures 11, 12a–12c, and 13). The structure beneath eastern Lake Issyk‐ Kul is similar to that of the basins imaged in this study. Some low‐Vp, low‐Vs and high‐s anomalies appear in the shallow crust (Figures 11, 13a, 13b, 13d, and 13f), while at great depths some high‐Vp, high‐Vs and low‐s anomalies show up (Figures 11 and 13). However, western Lake Issyk‐ Kul demonstrates structure similar to the central Tien Shan. The upper and middle crust shows low‐Vp, low‐Vs and high‐s anomalies, while in the lower crust and upper mantle some high‐Vp, high‐Vs and low‐s anomalies are illustrated (Figures 11, 12a–12c, and 13d–13f). These structural differences were not detected by previous local tomographic studies [e.g., Omuralieva et al., 2009] but they are clearly

observed in the present work. These results are supported by Pn tomography [e.g., Xu et al., 2007] and teleseismic tomography [e.g., Lei and Zhao, 2007]. Combining these results, it is inferred that the contrasts between eastern and western Lake Issyk‐Kul may extend from the crust down to the upper mantle. [17] It is also found that the Fergana fault shows a distinct boundary between central and western Tien Shan (Figures 11 and 13d–13f). In western Tien Shan, relative low‐Vp, low‐Vs and high‐s anomalies are revealed in the upper crust, while in the lower crust and upper mantle high‐Vp and high‐Vs anomalies are illustrated, and low‐s anomalies are confined to the lower crust (Figure 13f). However, in central Tien Shan, high‐Vp, high‐Vs and low‐s images are revealed in the upper crust, and low‐Vp, low‐Vs and high‐s anomalies appear in the lower crust and upper mantle (Figures 13d and 13f). These results suggest that the Fergana fault may

8 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 8. The same as Figure 7 but for S wave structure in map view.

represent a lithospheric‐scale boundary. In addition, a distinct low‐Vp and low‐Vs anomaly extending from the surface down to the upper mantle is clearly visible at the turning point of the Fergana fault from the NWN to NWW orientation (Figures 13d and 13e). Although a high‐s anomaly is not detected well in the upper crust, it evidently appears in the lower crust and upper mantle (Figure 13f). These results may indicate that the Fergana fault may provide a channel for the hot and wet material upwelling to the surface from the mantle. [18] Some high‐Vp, high‐Vs and low‐s anomalies are found in the collision zones between the Tien Shan and the Tarim basin in the south and Kazakh shield in the north (Figures 11, 12a, 12b, and 13a–13c). These high‐V and low‐s anomalies are tilted toward the Tien Shan along with the seismicity, supporting the notion of underthrusting of the Tarim and Kazakh lithosphere. Meanwhile, the other portions of these zones are imaged as prominent low‐V and high‐s anomalies (Figures 11 and 12c–12f), perhaps indi-

cating the intrusion of the hot and wet material into the crust from the mantle. 3.3. Synthetic Tests [19] To further check the vertical resolution of the tomographic images obtained, many synthetic tests were performed (Figure 14). The procedure of the synthetic test is the same as that of the checkerboard resolution test (Figures 7– 10). The only difference between them is in the input model. In the checkerboard test alternative positive and negative anomalies of up to ±5% are assigned to the 3‐D grid nodes in the input model, while in the synthetic test under the Tien Shan a high‐V anomaly of 5% above 35 km depth and a low‐V anomaly of up to −5% at depths of 35–80 km are put in the input model (Figures 14a and 14d). The output models (Figures 14b, 14c, 14e, and 14f) show a similar pattern of velocity anomalies as that in the input model (Figures 14a and 14d), though there exist some differences between them in details and there is some smearing in the

9 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 9. The same as Figure 7 but for the grid spacing is 1.0° × 1.0° in the horizontal directions.

horizontal directions and vertical direction in the output models, especially for the marginal regions of the model (Figures 14b, 14e, and 14f). These results all suggest that the main structural features, such as high‐V anomalies in the upper and middle crust and low‐V anomalies in the lower crust and upper mantle under central Tien Shan, obtained in the present study are reliable.

4. Discussion 4.1. Effect of the Moho Depth Variation [20] Many previous studies have shown that Moho depth variations can affect tomographic images of the crust and upper mantle [e.g., Zhao et al., 1992; Huang and Zhao, 2004; Lei and Zhao, 2006]. In the present study, to evaluate the effect of the large undulated (38–68 km) Moho topography under the Tien Shan (Figure 4b), I conducted one more inversion by considering a flat Moho discontinuity with an average depth of 50 km in the region (Figure 4a).

The obtained results (Figure 15) show a similar influence of the Moho topography on the images as inferred by previous studies [e.g., Zhao et al., 2005; Lei and Zhao, 2006]. The Moho topography does not change the broader pattern of velocity anomalies but the details do change (Figures 12a– 12c, 13a–13c, and 15), which could improve the understanding of dynamic processes of the Tien Shan mountain building. For example, to the west of Lake Issyk‐Kul the Vp and Vs models with the Moho topography show a continuous low‐V anomaly under the Kazakh shield from the mantle to the crust (Figures 12a and 12b), while those with a flat Moho discontinuity show a discontinuous low‐V anomaly (Figures 15a and 15b). To the east of Lake Issyk‐ Kul the Vp and Vs models with the Moho topography illustrate a high‐V anomaly in the middle and lower crust under the Kazakh shield (Figures 13a and 13b), while those without the Moho depth variations demonstrate a weaker high‐V anomaly (Figures 15d and 15e). In addition, under the Tien Shan the Vp and Vs models with an undulated

10 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 10. The same as Figure 9 but for S wave structure in map view.

Moho depth show a thicker high‐V layer in the upper and middle crust, and intermittent low‐V anomalies in the lower crust and upper mantle (Figures 12a, 12b, 13a, and 13b), but those with a flat Moho discontinuity show a thinner high‐V layer and a continuous low‐V anomaly (Figures 15a, 15b, 15d, and 15e). The sensitivity of tomographic images to the assumed Moho depth suggests that independent investigations to obtain a more accurate Moho depth are needed. 4.2. The Talas‐Fergana Fault [21] The Talas‐Fergana fault extends a distance of about 500 km from the Tarim basin in the south to the Kazakh shield in the north. The fault is a prominent right‐lateral strike‐slip fault and has accumulated at least 250 km offset since the Permian with a slip rate of 8–16 mm/yr [e.g., Burtman, 1975; Tapponnier and Molnar, 1979; Burtman et al., 1996], though there has been little seismic activity since last century. This is quite similar to the Red River fault in southwest China [e.g., Lei et al., 2009b].

[22] Previous studies demonstrated that the Talas‐Fergana fault is an important boundary between central and western Tien Shan. For example, tomographic images supported the presence of high‐V anomalies west of the fault [e.g., Roecker et al., 1993; Lei et al., 2002] and the existence of low‐V anomalies east of the Talas‐Fergana fault [e.g., Roecker et al., 1993; Ghose et al., 1998; Lei et al., 2002; Lei and Zhao, 2007; Xu et al., 2002; Omuralieva et al., 2009], which are generally consistent with the geochemical studies suggesting the wide presence of the basaltic rocks east of the Talas‐Fergana fault [Sobel and Arnaud, 2000]. However, different researchers showed different depth ranges for the fault. Shear wave splitting anisotropy results show that the fast orientations near the Talas‐Fergana fault are close to the E‐W direction. These are consistent with the anisotropy away from the fault in the Tien Shan but are in contrast to the NW‐SE orientation Talas‐Fergana fault (Figure 16b) [e.g., Makeyeva et al., 1992; Li and Chen, 2006; Jiang et al., 2010]. Seismic anisotropy could reflect deformation in the

11 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

Figure 11. Results of (a–f) Vp, (g–l) Vs, and (m–q) and s tomographic images in map view. Red colors denote low‐Vp, low‐Vs and high‐s anomalies, while blue colors indicate high‐Vp, high‐Vs and low‐s anomalies. The scale for Vp, Vs and s perturbations (in %) is shown to the right of Figure 11l. Depth values are shown on the left). Some traces denote major outlines in the region; see Figure 2. White circles show strong earthquakes (M > 6.0) that occurred from 1964 to 2008 in the region. These hypocenter parameters were reprocessed by Engdahl et al. [1998]. The scale for the magnitude of the earthquakes is shown on the right side of Figure 11l. 12 of 20

B09305

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

Figure 12. The same as Figure 11 but for vertical cross sections. Red colors denote low‐Vp, low‐Vs and high‐s anomalies, while blue colors indicate high‐Vp, high‐Vs and low‐s anomalies. The scale for Vp, Vs and s perturbations (in %) is shown at the bottom. Crosses denote smaller earthquakes (M < 6) that occurred within 25 km of the profile. White circles denote the larger earthquakes (M > 6) in a range of 25 km off the profile. The scale for both small and large earthquakes is shown at the bottom. These earthquakes were all extracted from the ISC (International Seismological Center) bulletins during 1964– 2004 that were reprocessed by Engdahl et al. [1998]. Dashed lines on the images denote the Moho discontinuity that was inferred from Vinnik et al. [2004] and Bassin et al. [2000]. Surface topography along the cross section is shown above each figure. No vertical exaggerations except for the surface topography. Locations of vertical cross sections are plotted in the insert map. The triangles in the inset denote the seismic stations used in the study. LIK, Lake Issyk‐Kul. 13 of 20

B09305

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 13. The same as Figure 12 but for two more vertical cross sections. FF, Talas‐Fergana fault; LIK, Lake Issyk‐Kul. lithosphere or mantle flow direction in the asthenosphere as a large result of lattice preferred orientation of anisotropic minerals, such as olivine and orthopyroxene, in the upper mantle. This may be used to suggest that the Talas‐Fergana

fault must be shallow and probably confined to the crust [Li and Chen, 2006]. [23] Receiver function analyses illustrated that the seismic velocity is 10% lower below 10 km depth east of the Talas‐ Fergana fault than that west of the fault, and suggested that

14 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 14. Results of a synthetic test for P and S wave velocity structures. (a and d) Input models. (b and e) Output models for P wave velocity structure. (c and f) Output models for S wave velocity structure. Red and blue colors denote low‐V and high‐V anomalies, respectively. The scale for the velocity structure is shown at the bottom. Dashed lines on the images denote the Moho discontinuity that was inferred from Vinnik et al. [2004] and Bassin et al. [2000]. Surface topography along the cross section is shown above each figure. No vertical exaggerations except for the surface topography. The inset map shows the locations of vertical cross sections and seismic stations used. LIK, Lake Issyk‐Kul. such a correlation may extend down to the upper mantle [Kosarev et al., 1993]. The present tomographic images clearly demonstrate that the structure differs east and west of the Talas‐Fergana fault extending from the crust to the upper mantle (Figures 11 and 13d–13f), which is generally consistent with previous tomographic results [Roecker et al., 1993], suggesting a lithospheric‐scale boundary. In addition to these differences between the eastern and western portions of the Tien Shan, the Talas‐Fergana fault may provide a channel for the upwelling of the hot and wet material from the mantle (Figures 11, 13d, and 13e). Such a structural feature is somewhat similar to the Red‐River fault in

southwest China and the Zhang‐Bo seismic zone in North China [e.g., Lei et al., 2008, 2009b]. 4.3. Causes of Large Earthquakes [24] In addition to small earthquakes, many large earthquakes occurred in the region since last century (Figures 1 and 16a). Source mechanism solutions (Figure 16a) and geological surveys all demonstrate that the ruptures of these earthquakes were dominated by thrusting, so the occurrence of large earthquakes could be closely related to the northward underthrusting of the Tarim lithosphere and the southward underthrusting of the Kazakh lithosphere [Lei

15 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 15. The same as Figures 12a–12c and 13a–13c but for a flat Moho discontinuity with an average depth of 50 km. and Zhao, 2007]. However, fluids from the hot, wet material rising from the mantle (Figures 11–13) may also have played an important role in the generation of these large earthquakes. [25] Focal depths of the large historic earthquakes in the region were not determined well due to lack of seismic data at that time (Figure 1), so it is hard to investigate the relationship between the velocity structure and distribution of these earthquakes. However, with the speedy development of global digital seismic stations and the accelerated accumulation of high‐quality arrival‐time data worldwide, source parameters, especially for focal depths, since 1964 have been significantly improved by using a large number of seismic phases with various distances [Engdahl et al., 1998]. I call the earthquakes after 1964 the “recent earthquakes.” It is evident that most of the recent large earthquakes occurred around 10–20 km depth, though few were

located around 30–35 km depth (Figures 11–13). It is seen that most of these recent large earthquakes were not only located in and around the collisional zone between the Pamir plateau and western Tien Shan, but also could be related to the pattern of velocity anomalies (Figures 11b and 11h). These large earthquakes mainly occurred in and around the zones with high‐Vp, high‐Vs and low‐s anomalies, but they are underlain by obvious low‐Vp, low‐Vs and high‐s anomalies (Figures 11–13). Such low‐V and high‐s anomalies could be closely related to elevated temperature in the Tien Shan [e.g., Akimoto, 1972; Karato, 1993]. This elevated temperature might generate partial melts or fluids that can increase the Poisson’s ratio and decrease the velocity. It is concluded that these low‐V and high‐s anomalies could be interpreted to be a fluid‐filled, fractured rock matrix that contributed to not only the formation of tectonic boundaries but also the initiation of these large

16 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

B09305

Figure 16. (a) The source mechanism solutions (M > 4.0) that were downloaded from the Web site of Harvard University (www. globalcmt.org/CMTsearch.html). (b) Shear wave splitting measurements obtained from previous studies [Makeyeva et al., 1992; Li and Chen, 2006; Jiang et al., 2010]. The orientation of a bar indicates fast orientations and the bar length is proportional to delay times. (c) GPS velocity field of the crust motion in the Tien Shan region relative to the stable Eurasian plate (http:// geoweb.mit.edu/∼tah/CAsia). earthquakes. The fluids may come from the hot and wet asthenospheric material or partial melt rising from the mantle. When the fluids enter the active faults, the effective normal stress across the fault planes would decrease and thus faulting could be triggered to generate these large

earthquakes in the Tien Shan region. A similar structure is also found in the source areas of the 1605 Qiongshan earthquake (M 7.5) in southernmost China [Lei et al., 2009a], the 1976 Tangshan earthquake (M 7.8) and 2006 Wen‐An earthquake (M 5.1) in North China [Huang and

17 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

Zhao, 2004; Lei et al., 2008, 2011], the 1995 Kobe earthquake (M 7.2) in Japan [Zhao et al., 1996], the 2001 Bhuj earthquake (M 7.6) in India [Mishra and Zhao, 2003], and the 2008 Wenchuan earthquake (M 8.0) in southwestern China [Lei and Zhao, 2009]. [26] Therefore, I conclude that the generation of these large earthquakes in the Tien Shan region may not only be associated with the northward underthrusting of the Tarim lithosphere and the southward underthrusting of the Kazakh lithosphere, but also could be related to the fluids that came from the hot and wet material rising from the mantle and that could decrease the effective normal stresses on the fault planes. 4.4. Structure Under the Basins [27] The present tomographic results reveal some significant structural features under some basins in and around Tien Shan. These features are very crucial to understand the geodynamic process of the mountain building. For example, under the Tarim and Fergana basins, some prominent low‐Vp, low‐Vs and high‐s anomalies are revealed at shallow depths and are underlain by high‐Vp, high‐Vs and low‐s anomalies at greater depths (Figures 11–13), which is consistent with those inferred by previous studies [e.g., Ghose et al., 1998; Kosarev et al., 1993; Roecker et al., 1993; Maceira and Ammon, 2009; Omuralieva et al., 2009], although the Tarim basin is located on the margins of the study region. These low‐V and high‐s anomalies under the basins at shallow depths may suggest the existence of the less‐compacted Cenozoic sediments due to erosion and denudation from the surrounding mountains [e.g., Ghose et al., 1993; Roecker et al., 1993; Maceira and Ammon, 2009; Omuralieva et al., 2009; Zheng et al., 2010], while high‐V and low‐s anomalies under the basins at greater depths may suggest the existence of craton‐like strong lithospheric material. Such a structural feature under the Tarim basin may support the fact that the stress can be effectively transferred to the Tien Shan orogenic belt from the India‐Asia collision front [e.g., England and Houseman, 1985]. A similar seismic structure is also found under other basins in China, such as the Dzugaria and Sichuan basins, by using surface wave and Pn tomography [Friederich, 2003; Liang et al., 2004; H. Li et al., 2009; Zheng et al., 2010]. 4.5. The Tien Shan Orogenic Belt [28] A variety of geological surveys and geophysical observations have been conducted to unravel the mechanism of the Tien Shan orogenic belt, but it still remains an enigma to us. Some researchers suggested that the mountain building is due to shortening in the crust accompanied by similar movement in the lithosphere [e.g., Fleitout and Froidevaux, 1982; Yin et al., 1998]. Such shortening could amount to 20 mm/yr (Figure 16c) from the India‐Asia collision transferring stress through the strong Tarim lithosphere [e.g., Tapponnier et al., 1981, 1990; England and Houseman, 1985, 1986; Abdrakhmatov et al., 1996; Yin and Harrison, 2000; Thompson et al., 2002]. Others suggested that the Tien Shan orogenic belt was caused by the underthrusting of the Tarim and Kazakh lithosphere, which has been demonstrated by seismic tomography [e.g., Roecker et al., 1993; Lei and Zhao, 2007], source mechanism solutions [e.g., Ni,

B09305

1978; Nelson et al., 1987], and receiver function analyses [e.g., Chen et al., 1997]. [29] In the present study some high‐Vp, high‐Vs and low‐s anomalies exist around the collision zones between the Tien Shan mountains and the Tarim basin in the south and the Kazakh shield in the north. The distribution of both small and large earthquakes is somewhat coupled with these high‐V and low‐s zones that are tilted toward the Tien Shan mountains with depth (Figures 12a–12d and 13). These results are generally consistent with previous studies [e.g., Roecker et al., 1993; Xu et al., 2002; Lei and Zhao, 2007], suggesting the northward underthrusting of the Tarim lithosphere and the southward underthrusting of the Kazakh lithosphere under the Tien Shan mountain. On the other hand, the present high‐resolution models also reveal some low‐Vp, low‐Vs and high‐s anomalies in other portions of these collision zones (Figures 11 and 12c–12f), perhaps indicating that the hot and wet material has been intruded into the crust from the mantle. This is quite similar to previous results [e.g., Omuralieva et al., 2009]. [30] The structure beneath central Tien Shan is dominated by high‐Vp, high‐Vs, and low‐s anomalies in the shallow crust, possibly suggesting the existence of the Paleozoic crystalline basement rocks, while low‐Vp, low‐Vs and high‐s anomalies are visible in the deep crust and upper mantle, perhaps indicating the upwelling of the hot and wet material from the mantle (Figures 11–13). These features have been demonstrated by magnetotelluric soundings that show lower resistivity under central Tien Shan [e.g., Bielinski et al., 2003] and are generally consistent with previous seismic studies [e.g., Kosarev et al., 1993; Roecker et al., 1993; Ghose et al., 1998; Xu et al., 2002; Vinnik et al., 2004; Kumar et al., 2005; Z. Li et al., 2009; Omuralieva et al., 2009]. However, there exist some differences in the location and depth range of the low‐V anomalies under the Tien Shan. Roecker et al. [1993] integrated local with teleseismic data in central and west Tien Shan and they concluded that the low‐V anomaly appears under central Tien Shan and extends down to 150 km depth but not below 300 km depth. The teleseismic tomography of Lei and Zhao [2007] revealed that the low‐V anomalies under the Tien Shan may extend down to the mantle transition zone under the Tarim basin and down to its top under the Kazakh shield, respectively. The receiver function analyses revealed that the mantle transition zone northwest of Lake Issyk‐Kul becomes thinner, suggesting the upwelling of the hot and wet material beneath the Kazakh shield from the lower mantle [Tian et al., 2010]. The present local tomographic models further confirm that prominent low‐Vp, low‐Vs and high‐s anomalies indeed exist in the lower crust and upper mantle of the Tien Shan (Figures 11–13). Some researchers suggested that the low‐V and high‐s anomalies may be a small plume [e.g., Sobel and Arnaud, 2000; Friederich, 2003; Vinnik et al., 2004], while others indicated that they may be a branch of the small‐scale mantle convection [e.g., Roecker et al., 1993; Wolfe and Vernon, 1998; Tian et al., 2010]. Either way, these low‐V and high‐s anomalies under the Tien Shan may suggest the upwelling of the hot and wet material from the mantle. Some detailed differences between low‐V anomalies from different researchers may suggest that further investigations should still be carried out to obtain a

18 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

more reliable seismic structure under the Tien Shan so as to better understand the geodynamic process of the mountain building. [31] Thus it is inferred that the Tien Shan mountain building could be closely related to not only the underthrusting of the Tarim and Kazakh lithosphere under the Tien Shan, but also associated with the upwelling of the hot material from the mantle, as proposed by previous studies [e.g., Roecker et al., 1993; Lei and Zhao, 2007].

5. Conclusions [32] A large number of high‐quality P and S wave arrival time data are inverted to determine 3‐D tomographic images of Vp, Vs and s under central and western Tien Shan. The results show that prominent high‐Vp, high‐Vs and low‐s anomalies revealed at shallow depths under the Tien Shan mountains are underlain by remarkable low‐Vp, low‐Vs and high‐s anomalies extending down to the mantle. In the collision zones between the Tien Shan and the Tarim basin in the south and the Kazakh shield in the north, some high‐ Vp, high‐Vs and low‐s anomalies that are tilted toward the Tien Shan with depth have been revealed and some earthquakes occurred along these zones. On the other hand, the other parts of these zones are imaged as low‐Vp, low‐Vs and high‐s anomalies, suggesting the intrusion of the hot and wet material into the crust from the mantle. These results all suggest that both the underthrusting of the Tarim and Kazakh lithosphere and the upwelling of the hot and wet material from the mantle may have played an important role in the mountain building. [33] The results illustrate that under some basins, such as the Tarim basin and the Fergana basin, some low‐Vp, low‐ Vs and high‐s anomalies appear at shallow depths from Cenozoic sediments. At greater depths high‐Vp, high‐Vs and low‐s anomalies dominate, perhaps indicating the relatively strong craton‐like structure that can help transfer the stress from the Inda‐Asia collision. In addition, a prominent velocity contrast extending down to the upper mantle is observed between east and west of the Talas‐Fergana fault. Meanwhile, a remarkable low‐Vp and low‐Vs column is visible at the turning point of the Talas‐Fergana fault from NWN in the south to NWW in the north. These results suggest that the fault may be a lithospheric‐scale boundary and might provide a channel for the upwelling of the hot and wet material from the mantle to the surface. [34] Acknowledgments. The author is grateful to Dapeng Zhao for providing the tomographic code used in the present study, to the IRIS Manager Center for providing the waveform data and R. Engdahl for providing the ISC arrival time data, to G. Zhang during the data process, to L. Vinnik for providing the Moho discontinuity model, and to D. Zhao and S. Cramp for thoughtful discussions. This work was partially supported by National Natural Science Foundation of China (40774044 and 40974021) and sponsored by the Scientific Research Foundation for Returned Oversea Chinese Scholars, State Education Ministry, the research grant (A07086) from the Chinese Earthquake Study Foundation, and Beijing Natural Science Foundation (8092028) to J. Lei. The GMT software package distributed by Wessel and Smith [1995] was used for plotting the figures. The author appreciates R. Nowack, the Associate Editor, and two anonymous referees for providing many constructive comments and suggestions, which improved the manuscript.

B09305

References Abdrakhmatov, K., et al. (1996), Relatively recent construction of the Tien Shan inferred from GPS measurements of present‐day crustal deformation rates, Nature, 384, 450–453, doi:10.1038/384450a0. Akimoto, I. (1972), The system MgO‐FeO‐SiO2 at high pressures and temperatures—Phase equilibria and elastic properties, Tectonophysics, 13, 161–187, doi:10.1016/0040-1951(72)90019-4. Bassin, C., G. Laske, and G. Masters (2000), The current limits of resolution for surface wave tomography in North America, Eos Trans. AGU, 81(48), Fall Meet. Suppl., Abstract S12A‐03. Bielinski, R., S. Park, A. Rybin, V. Batalev, S. Jun, and C. Sears (2003), Lithospheric heterogeneity in the Kyrgyz Tien Shan imaged by manetotelluric studies, Geophys. Res. Lett., 30(15), 1806, doi:10.1029/ 2003GL017455. Burtman, V. (1975), Structural geology of the Variscan Tien Shan, Am. J. Sci., 275A, 157–186. Burtman, V., S. Skobele, and P. Molnar (1996), Late Cenozoic slip on the Talas‐Fergana fault, the Tien Shan, central Asia, Geol. Soc. Am. Bull., 108(8), 1004–1021, doi:10.1130/0016-7606(1996)1082.3.CO;2. Chen, C., H. Lu, D. Jia, D. Cai, and S. Wu (1999), Closing history of the southern Tianshan oceanic basin, western China: An oblique collisional orogeny, Tectonophysics, 302, 23–40, doi:10.1016/S0040-1951(98) 00273-X. Chen, Y., S. Roecker, and G. Kosarev (1997), Elevation of the 410 km discontinuity beneath the central Tien Shan: Evidence for a detached lithospheric root, Geophys. Res. Lett., 24, 1531–1534, doi:10.1029/ 97GL01434. Engdahl, R., R. van der Hilst, and R. Buland (1998), Global teleseismic earthquake relocation with improved travel times and procedures for depth determination, Bull. Seismol. Soc. Am., 88, 722–743. England, P., and G. Houseman (1985), The influence of lithospheric strength heterogeneities on the tectonics of Tibet and surrounding regions, Nature, 315, 297–301, doi:10.1038/315297a0. England, P., and G. Houseman (1986), Finite strain calculations of continental deformation: 2. Comparison with the India‐Asia collision zone, J. Geophys. Res., 91, 3664–3676, doi:10.1029/JB091iB03p03664. Fleitout, L., and C. Froidevaux (1982), Tectonics and topogaphy for a lithosphere containing density heterogeneities, Tectonics, 1, 21–56, doi:10.1029/TC001i001p00021. Friederich, W. (2003), The S‐velocity structure of the East Asian mantle from inversion of shear and surface waveforms, Geophys. J. Int., 153, 88–102, doi:10.1046/j.1365-246X.2003.01869.x. Ghose, S., M. Hamburger, and J. Virieux (1998), Three‐dimensional velocity structure and earthquake locations beneath the northern Tien Shan of Kyrgyzstan, central Asia, J. Geophys. Res., 103, 2725–2748, doi:10.1029/97JB01798. Huang, J., and D. Zhao (2004), Crustal heterogeneity and seismotectonics of the region around Beijing, China, Tectonophysics, 385, 159–180, doi:10.1016/j.tecto.2004.04.024. Jiang, L., Y. Li, and Q. Wu (2010), The shear wave splitting of the central Tien Shan and its implications, Chin. J. Geophys., 53(6), 1399–1408. Karato, S. (1993), Importance of anelasticity in the interpretation of seismic tomography, Geophys. Res. Lett., 20, 1623–1626, doi:10.1029/ 93GL01767. Kennett, B. L. N., and R. Engdahl (1991), Travel times for global earthquake location and phase identification, Geophys. J. Int., 105, 429–465, doi:10.1111/j.1365-246X.1991.tb06724.x. Kondorskaya, V., and N. Shebalin (1982), New Catalogs of Strong Earthquakes in the USSR From Ancient Times Through 1977, 608 pp., World Data Cent. for Solid‐Earth Geophys., Boulder, Colo. Kosarev, G., N. Petersen, and L. Vinnik (1993), Receiver functions for the Tien Shan analog broadband network: Contrasts in the evolution of structures across the Talasso‐Fergana fault, J. Geophys. Res., 98, 4437–4448, doi:10.1029/92JB02651. Kumar, P., X. Yuan, R. Kind, and G. Kosarev (2005), The lithosphere‐ asthenosphere boundary in the Tien Shan‐Karakoram region from S receiver function: Evidence for continental subduction, Geophys. Res. Lett., 32, L07305, doi:10.1029/2004GL022291. Lei, J., and D. Zhao (2005), P wave tomography and origin of the Changbai intraplate volcano in northeast Asia, Tectonophysics, 397, 281–295, doi:10.1016/j.tecto.2004.12.009. Lei, J., and D. Zhao (2006), Global P wave tomography: On the effect of various mantle and core phases, Phys. Earth Planet. Inter., 154, 44–69, doi:10.1016/j.pepi.2005.09.001. Lei, J., and D. Zhao (2007), Teleseismic P wave tomography and the upper mantle structure of the central Tien Shan orogenic belt, Phys. Earth Planet. Inter., 162, 165–185, doi:10.1016/j.pepi.2007.04.010.

19 of 20

B09305

LEI: SEISMIC HETEROGENEITY OF THE TIEN SHAN

Lei, J., and D. Zhao (2009), Structural heterogeneity of the Longmenshan fault zone and the mechanism of the 2008 Wenchuan earthquake (M 8.0), Geochem. Geophys. Geosyst., 10, Q10010, doi:10.1029/2009GC002590. Lei, J., H. Zhou, and D. Zhao (2002), 3‐D velocity structure of P wave in the crust and upper mantle beneath the Pamir and adjacent region, Chin. J. Geophys., 45(6), 802–811. Lei, J., F. Xie, C. Lan, C. Xing, and S. Ma (2008), Seismic images under the Beijing region inferred from P and PmP data, Phys. Earth Planet. Inter., 168, 134–146, doi:10.1016/j.pepi.2008.06.005. Lei, J., D. Zhao, B. Steinberger, B. Wu, F. Shen, and Z. Li (2009a), New seismic constraints on the upper mantle structure of the Hainan plume, Phys. Earth Planet. Inter., 173, 33–50, doi:10.1016/j.pepi.2008.10.013. Lei, J., D. Zhao, and Y. Su (2009b), Insight into the origin of the Tengchong intraplate volcano and seismotectonics in southwest China from local and teleseismic data, J. Geophys. Res., 114, B05302, doi:10.1029/2008JB005881. Lei, J., D. Zhao, F. Xie, and J. Liu (2011), An attempt to detect temporal variations of crustal structure in the source area of the 2006 Wen‐An earthquake in North China, J. Asian Earth Sci., 40, 958–976, doi:10.1016/j.jseaes.2010.07.003. Li, A., and C. Chen (2006), Shear wave splitting beneath the central Tien Shan and tectonic implications, Geophys. Res. Lett., 33, L22303, doi:10.1029/2006GL027717. Li, H., W. Su, C. Wang, and Z. Huang (2009), Ambient noise tomography in western Sichuan and eastern Tibet, Earth Planet. Sci. Lett., 282, 201–211, doi:10.1016/j.epsl.2009.03.021. Li, Z., S. Roecker, Z. Li, B. Wei, H. Wang, G. Schelochkov, and V. Bragin (2009), Tomographic images of the crust and upper mantle beneath the western Tien Shan from the MANAS broadband deployment: Possible evidence for lithospheric delamination, Tectonophysics, 477, 49–57, doi:10.1016/j.tecto.2009.05.007. Liang, C., X. Song, and J. Huang (2004), Tomographic inversion of Pn Travel times in China, J. Geophys. Res., 109, B11304, doi:10.1029/ 2003JB002789. Maceira, M., and C. Ammon (2009), Joint inversion of surface wave velocity and gravity observations and its application to central Asian basins shear velocity structure, J. Geophys. Res., 114, B02314, doi:10.1029/ 2007JB005157. Makeyeva, L., L. Vinnik, and S. Roecker (1992), Shear‐wave splitting and small‐scale convection in the continental upper mantle, Nature, 358, 144–147, doi:10.1038/358144a0. Mishra, O. P., and D. Zhao (2003), Crack density, saturation rate and porosity at the 2001 Bhuj, India, earthquake hypocenter: A fluid‐driven earthquake?, Earth Planet. Sci. Lett., 212, 393–405, doi:10.1016/S0012821X(03)00285-1. Molnar, P., and S. Ghose (2000), Seismic moments of major earthquakes and the rate of shortening across the Tien Shan, Geophys. Res. Lett., 27(16), 2377–2380, doi:10.1029/2000GL011637. Mooney, W., G. Laske, and G. Masters (1998), CRUST 5.1: A global crustal model at 5 × 5 degrees, J. Geophys. Res., 103, 727–747, doi:10.1029/97JB02122. Nelson, M., R. McCaffrey, and P. Molnar (1987), Source parameters for 11 earthquakes in the Tien Shan, central Asia, determined by P and Sh waveform inversion, J. Geophys. Res., 92, 12,629–12,648, doi:10.1029/JB092iB12p12629. Ni, J. (1978), Contemporary tectonics in the Tien Shan region, Earth Planet. Sci. Lett., 41, 347–354, doi:10.1016/0012-821X(78)90189-9. Omuralieva, A., J. Nakajima, and A. Hasegawa (2009), Three‐dimensional seismic velocity structure of the crust beneath the central Tien Shan, Kyrgyzstan: Implications for large‐ and small‐scale mountain building, Tectonophysics, 465, 30–44, doi:10.1016/j.tecto.2008.10.010. Paige, C., and M. Saunders (1982), LSQR: An algorithm for sparse linear equations and sparse least squares, Assoc. Comput. Math. Trans. Math. Software, 8, 43–71, doi:10.1145/355984.355989. Poupinet, G., et al. (2002), Intracontinental subduction and Palaeozoic inheritance of the lithosphere suggested by a teleseismic experiment across the Chinese Tien Shan, Terra Nova, 14(1), 18–24, doi:10.1046/ j.1365-3121.2002.00391.x. Roecker, S. (2001), Constraints on the crust and upper mantle of the Kyrgyz Tien Shan from the preliminary analysis of GHENGIS broadband data, Russ. Geol. Geophys., 42, 1554–1565.

B09305

Roecker, S., T. Sabitova, L. Vinnik, Y. Burmakov, M. Golvanov, R. Mamatkanova, and L. Munirova (1993), Three‐dimensional elastic wave velocity structure of the western and central Tien Shan, J. Geophys. Res., 98, 15,779–15,795, doi:10.1029/93JB01560. Sobel, E., and N. Arnaud (2000), Cretaceous‐Paleogene basaltic rocks of the Tuyon basin, NW China and the Kyrgz Tian Shan: The trace of a small plume, Lithos, 50, 191–215, doi:10.1016/S0024-4937(99)00046-8. Sobel, E., and T. Dumitru (1997), Thrusting and exhumation around the margins of the western Tarim basin during the India‐Asia collision, J. Geophys. Res., 102, 5043–5063, doi:10.1029/96JB03267. Tapponnier, P., and P. Molnar (1979), Active faulting and Cenozoic tectonics of the Tien Shan, Mongolia, and Baikal regions, J. Geophys. Res., 84, 3425–3459, doi:10.1029/JB084iB07p03425. Tapponnier, P., et al. (1981), The Tibetan side of the India‐Eurasian collision, Nature, 294, 405–410, doi:10.1038/294405a0. Tapponnier, P., R. Lacassion, P. Leloup, U. Scharer, D. Zhong, H. Wu, X. Liu, S. Ji, L. Zhang, and J. Zhong (1990), The Ailao Shan/Red‐River metamorphic belt: Tertiary left‐lateral shear between Indochina and South China, Nature, 343, 431–437, doi:10.1038/343431a0. Thompson, S., R. Weldon, C. Rubin, K. Abdrakhmatov, P. Molnar, and G. Berger (2002), Late Quaternary slip rates across the central Tien Shan, Kyrgyzstan, central Asia, J. Geophys. Res., 107(B9), 2203, doi:10.1029/ 2001JB000596. Tian, X., D. Zhao, H. Zhang, Y. Tian, and Z. Zhang (2010), Mantle transition zone topography and structure beneath the central Tien Shan orogenic belt, J. Geophys. Res., 115, B10308, doi:10.1029/2008JB006229. Vinnik, L., C. Reigber, I. Aleshin, G. Kosarev, M. Kaban, S. Oreshin, and S. Roecker (2004), Receiver function tomography of the central Tien Shan, Earth Planet. Sci. Lett., 225, 131–146, doi:10.1016/j. epsl.2004.05.039. Wessel, P., and W. Smith (1995), New version of the Generic Mapping Tools released, Eos Trans. AGU, 76(33), 329, doi:10.1029/95EO00198. Wolfe, C., and F. Vernon (1998), Shear‐wave splitting at central Tien Shan: Evidence for rapid variation of anisotropic patterns, Geophys. Res. Lett., 25, 1217–1220, doi:10.1029/98GL00838. Xu, Y., F. Liu, J. Liu, and H. Chen (2002), Crust and upper mantle structure beneath western China from P wave travel time tomography, J. Geophys. Res., 107(B10), 2220, doi:10.1029/2001JB000402. Xu, Y., Z. Li, and S. Roecker (2007), Uppermost mantle structure and its relation with seismic activity in the central Tien Shan, Geophys. Res. Lett., 34, L10304, doi:10.1029/2007GL029708. Yin, A., and T. Harrison (2000), Geological evolution of the Himalayan‐ Tibetan orogen, Annu. Rev. Earth Planet. Sci., 28, 211–280, doi:10.1146/annurev.earth.28.1.211. Yin, A., S. Nie, P. Craig, T. Harrison, F. Ryerson, X. Qian, and G. Yang (1998), Late Cenozoic tectonic evolution of the southern Chinese Tian Shan, Tectonics, 17, 1–27, doi:10.1029/97TC03140. Zhao, D. (2004), Global tomographic images of mantle plumes and subducting slabs: Insight into deep Earth dynamics, Phys. Earth Planet. Inter., 146, 3–34, doi:10.1016/j.pepi.2003.07.032. Zhao, D., A. Hasegawa, and S. Horiuchi (1992), Tomographic imaging of P and S wave velocity structure beneath northeastern Japan, J. Geophys. Res., 97, 19,909–19,928, doi:10.1029/92JB00603. Zhao, D., H. Kanamori, H. Negishi, and D. Wiens (1996), Tomography of the source area of the 1995 Kobe earthquake: Evidence for fluids at the hypocenter?, Science, 274, 1891–1894, doi:10.1126/science.274.5294.1891. Zhao, D., S. Todo, and J. Lei (2005), Local earthquake reflection tomography of the Landers aftershock area, Earth Planet. Sci. Lett., 235, 623–631, doi:10.1016/j.epsl.2005.04.039. Zhao, J., G. Liu, Z. Lu, X. Zhang, and G. Zhao (2003), Lithospheric structure and dynamic processes of the Tianshan orogenic belt and the Junggar basin, Tectonophysics, 376, 199–239, doi:10.1016/j.tecto. 2003.07.001. Zheng, X., W. Jiao, C. Zhang, and L. Wang (2010), Short‐period Rayleigh‐ wave group velocity tomography through ambient noise cross‐correlation in Xinjiang, northwest China, Bull. Seismol. Soc. Am., 100(3), 1350–1355, doi:10.1785/0120090225. J. Lei, Seismological Laboratory, Institute of Crustal Dynamics, China Earthquake Administration, Beijing 100085, China. ([email protected] com; [email protected]‐icd.cn)

20 of 20

Suggest Documents