Yunnan is an ideal location for investigating continental geodynamics, earthquake environment gestating and earthquake prediction. Since the 1980s, much geophysical research has been carried out in Yunnan, for instance, seismic sounding in 1982, 1987 and 1999, and inversion of gravity data (He Xiang et al., 1986), which shows the major characteristics of changes of the crust thickness in Yunnan.
The northwest of Yunnan is located on the southeast boundary of Qinghai-Tibetan plateau, in the middle of Hengduan Mountains or the Three Rivers and its basin, between the belts of the southeast arc turning point in Tethys-Himalaya tectonics and on the orogenic belt of Gondwana land and the Eurasian Plate, connecting the Yantze Platform and the Western Yunnan geosynclinal fold belt. It is one of the most frequent and intense earthquake areas. The northwest Yunnan consists of the three rivers block, Garzê block, central Yunnan block and northwest Yunnan rhombus block (Heqing-Dali rhombus block) (Institute of Geology, SSB, 1990). The Binchuan area (100.50°-100.70°E, 25.6°-26.00°N), located between the Honghe Fault zone and Chenghai fault zone, belongs to the northwest Yunnan rhombus block. The area is controlled mainly by the Chenghai Fault, which has been a significant active fault zone since the Quaternary period with M≥7.0 earthquakes happening at both of its ends, while in the Binchuan area's south-central zone, there has not been any earthquake of M≥6.0 in 200 years, so the study of the Binchuan region's crustal structure is significant.
With the development of digital seismic observation technology, it has become relatively easy to study the Earth's internal structure by using actual seismic waveforms. The teleseismic receiver function is a mature method to study crustal structure and composition which has brought a lot of research results. Yuan et al. (1997) successfully used the receiver function stacking methods to study the lateral variation of upper mantle discontinuity to obtain the mantle structure of the Qinghai-Tibetan plateau and Andean region. Duiker initially proposed the receiver function CMP stacking method. Liu Qiyuan et al.(1996, 1999, 2000) used broadband seismic array observations and the receiver function method to carry out a lot of research, such as receiver functions of nonlinear inversion and synthesis of three-dimensional lateral inhomogeneous medium teleseismic body waves. Wu Qingju et al. (1998) has adopted and improved the time-domain deconvolution technique to study the crustal structure of Qinghai-Tibetan Plateau. Zou Zuihong et al. (2003) has drawn a conclusion on the advantages and disadvantages of the SV-component receiver function: the receiver function's amplitude is stable as its epicentral distance changes, the waveform is simple, and the PS converted wave outstanding, with a better convergence in inversion.
Hu Jiafu et al. (2003) used 3-component teleseismic 16-bite digital broad-band seismic records and gathered the wave receiver function values from 23 seismic stations to calculate the crust thickness, the velocity structure of S-wave and spatial distribution of Poisson's ratio. Wu Jianping et al. (2001) applied inversion of the broadband teleseismic receiver function from the digital seismic stations in Yunnan to obtain the S-wave's velocity structures at the depth from 0-100km. Zha Xiaohui et al. (2013) applied H-K stacking of receiver functions to find the crust thickness and V_{P}/V_{S} ratio. He Chuansong et al. (2004a) and Feng Jing et al. (2012) applied receiver functions to study the S-wave's velocity structure of the Tengchong region. Furthermore, He Chuansong et al. (2012b) studied the deep structure of the northwest Yunnan as well.
Since the implementation of Active Source Project and the Himalaya Array, many seismic stations were set up in recent years in the northwest Yunnan region, which provides a better platform for seismology research. To carry out earthquake research in the Binchuan area with the monitoring data and information from the seismic stations is both meaningful and fundamental. The Binchuan region, from 100.50°-100.70°E, 25.6°-26.00°N, with an area of 40km×45km, located between the Honghe fault zone and Chenghai fault zone, belongs to the northwest Yunnan rhombus block. There are 15 short-period seismographs and 1 broadband seismograph installed in the Binchuan region, with a distance less than 15km between each seismograph. The fine crust structure research is beneficial from the high density in seismograph and the unique geography in Binchuan, so it trends to receive a precise result on crust thickness from teleseismic receiver functions data. This paper aims to calculate the precise crust thickness and the distribution of Poisson's ratio in Binchuan by using body-wave receiver functions gathered from the data and information of 16 seismic stations in this region.
1 DATA SELECTIONIn 2011, the Institute of Geophysics of China Earthquake Administration and Earthquake Administration of Yunnan Province developed a large-volume air-gun source launcher for conventional observation (The Binchuan Active Source Transmitter), at the Dayindian Reservoir in Binchuan, Western Yunnan. It is the first in the world and is mainly used for experiments for detecting active sources and to serve as an experiment platform to practise earthquake prediction by using new techniques and methods, observing the underground structure as the radar does in the sky. Researchers abroad, such as the San Andreas Fault Observatory at Depth (SAFOD), use the drillhole active sources to monitor the wave velocity changes prior to earthquakes. There are 34 seism stations in the Binchuan Active Source Observation Program. This paper uses the telegraphic waveform data from 15 short-period stations and one broadband station in the Binchuan region (100.50°-100.70°E, 25.6°-26.00°N), within a range of 40km×45km. Each of the stations is different from the others and complicated in geography, so the differences in crustal components and crust structure are likely to be found in the receiving functions. Fig. 1 shows the distribution of the stations and the spatial distribution of the teleseisms in the research region. Due to deficiency of earthquake data from the north and southwest area, the earthquake data in this research is mainly from the stations in the northeast, the east, the southeast, the south, the northwest and the west (anti-azimuth 31°-300°). On the Richter Scale, the earthquakes chosen in this research include one of magnitude 8.0 or above, 13 of magnitude between 7.0 and 7.9, 123 of magnitude between 6.0 and 6.9, as well as 101 of magnitude between 5.8 and 5.9.
A teleseismic P-wave arrived at the stations with a relatively strong phase velocity. As marked on the three-component record, the vertical components are mainly dominated by the P-wave, and the horizontal component is the S-wave. The teleseismic records show that a series of scattered waves followed the wave P, in which the Ps converted wave phases are especially outstanding. The series of scattered waves in the horizontal component is called the station receiving function. The receiving function, which shows the distribution tendency of P-wave, Ps converted wave, and multi-reflection wave (Fig. 2) when the earthquake wave arrives beneath the station and spreads upwards within the crust layers, is a very important way to study the mantle structure of the crust. Studying the vertical components of teleseismic observation data as a hypocenter or primary wave propagation path and making deconvolution in both the radial and the tangential component is able to remove hypocenter time function, the effect of hypocenter or propagation effect, and get the Ps converted wave and multi-reflection wave beneath the station. The receiving function is used to separate the Ps converted wave from the complicated earthquake data, and the Ps converted wave enables us to study the structure of the crust and the upper mantle.
If the Moho layer is the deepest reflection interface (Fig. 3), the time difference and the crust thickness of the Ps converted wave and the direct arriving P-wave can be calculated as the formula (Zandt G. et al., 1995):
$H = \frac{{{t_{{\rm{Ps}}}} - {t_{\rm{P}}}}}{{\sqrt {\frac{1}{{V_{\rm{S}}^2}} - {p^2}} - \sqrt {\frac{1}{{V_{\rm{P}}^2}} - {p^2}} }}$ | (1) |
where p is ray parameters, H is the crust thickness, V_{P} and V_{S} area respectively the P-wave velocity and the shear wave velocity. Therefore, similarly the formula between the time difference and the crust thickness of the multiple reflections Pp, Ps and the Ps converted wave is as follows:
$H = \frac{{{t_{{\rm{PpPs}}}} - {t_{{\rm{Ps}}}}}}{{2\sqrt {\frac{1}{{V_{\rm{P}}^2}} - {p^2}} }}$ | (2) |
The V_{P}/V_{S} ratio can be calculated as follows:
$\frac{{{V_{\rm{P}}}}}{{{V_{\rm{S}}}}} = {\left\{ {\left( {1 - {p^2}V_{\rm{P}}^2} \right){{\left[ {2\left( {\frac{{{t_{Ps}} - {t_{\rm{P}}}}}{{{t_{{\rm{PpPs}}}} - {t_{{\rm{Ps}}}}}}} \right) + 1} \right]}^2} + {p^2}V_{\rm{P}}^2} \right\}^{1/2}}$ | (3) |
Though the right side of the formula contains the P-wave velocity, it has a faint effect on the value of V_{P}. Since the average P-wave velocity of the crust is 6.00-6.75 km/s, calculating in a given period of time, the difference between the maximum velocity and the minimum is only 0.05, which enables us to calculate the change of Poisson's ratio, which is ≤0.02. Once we get the wave velocity ratio, we can get the value of Poisson's ratio σ according to the theory of elastic mechanics as below:
$\begin{array}{l} \frac{{{V_{\rm{P}}}}}{{{V_{\rm{S}}}}} = {\left[ {\frac{{2\left( {1 - \sigma } \right)}}{{1 - 2\sigma }}} \right]^{\frac{1}{2}}}\\ \sigma = \frac{{1 - \frac{1}{2}{{\left( {\frac{{{V_{\rm{P}}}}}{{{V_{\rm{S}}}}}} \right)}^2}}}{{1 - {{\left( {\frac{{{V_{\rm{P}}}}}{{{V_{\rm{S}}}}}} \right)}^2}}} \end{array}$ | (4) |
In order to calculate a relatively precise value of the crust thickness, the wave velocity ratio, and the Poisson's ratio, we have to choose receiver functions in various options from each station. The initial selection is based on the effect of final iteration. Generally, the iterative residue is at least less than 15%. Then, among the receiving functions chosen, only those of a clear phase in its Ps converted wave and PpPs reflecting wave are selected for seismic phase identification and reading of corresponding time delays.
Therefore, though there are 238 teleseismic records from all the stations, the receiving functions actually taken in the research are fewer, owing to the interference factor and different Moho interface levels from each station. There is a big different in the availability of receiving functions in each station which doesn't provide a lot of receiving functions of good relevance and morphology. However, we have at least 25 receiving functions in all stations, among which there are more than 35 receiving functions from 13 stations. Therefore, there is an average statistical effect. Considering that only a single receiving function is not objective to detect the specific location of multiple reflections due to the level of instrument stability, correlated receiving functions clustering is used to distinguish the multiple reflections (Fig. 3).
Table 1 shows the analysis results of receiving functions from the stations in the Binchuan area. The value of H_{1} is calculated from formula (1) about the time difference of the average P-wave velocity and the average S-wave velocity and direct arriving P-wave and Ps converted wave, while the value of H_{2} is obtained from the time difference between the average P-wave velocity (which is shown directly in the monitoring data) and the receiving function (formula (2)). Theoretically, if the interface of scanning width of the receiver is in the same level, then H_{1} should be equal to H_{2}. However, the obtained result doesn't follow that, and the range difference between the two values varies a lot from the stations. The main reason is that the S-wave velocity applied is not accurate enough (or else possibly owing to the sloped interface), which finally caused an inaccurate H_{1} value. Therefore, the H_{2} value, calculated from the time difference between the average P-wave velocity (which is shown directly in the monitoring data) and the receiving function, is taken as the result of the receiving function in this article. The value H_{1} is displayed to make a comparison with the value H_{1.} In this research, station 53258 is far away from the others so its interpolation results are less reliable, but its station data is still reliable because the number of receiving functions selected from this station is 85, which shows great credibility in its data.
From Table 1, it can be seen that there is a little difference between the two crust thickness values calculated respectively from formula (1) and formula (2) and the crust thicknesses of each station are close to each other, with a difference of less than 2km. However, it's not the case at Paiyingxi station (53258), where the value H_{1} is 49.7km and H_{2} is 42.7km, with a difference of 7km. The calculated average crust thickness of the Binchuan area is: H_{1} is 46.2km and H_{2} is 45.3km, and the value H_{2}=45.3km is basically consistent with the result of artificial seismic sounding obtained by Kan Rongju et al. (1986).
The distribution of the crust thickness of Binchuan area is quite different from north to south (Fig. 4), the thickness is gradually reduced from the north to the south which confirms the result of deep in the north and shallow in the south from the research carried out by Kan Rongju et al. (1986), Wu Jianping et al. (2001), Hu Jiafu et al. (2003), Li Yonghua et al. (2009), Zhang Xiaoman et al. (2011). From Paiying (42.1km) in the south to Beihe village (46.6km) in the north, there is an increase of 4.5km in thickness, showing a significant change in the south and the north. Except for that, this paper has found that the change of lateral crust thickness in the Binchuan area is obvious: from the west to the east the thickness gradually increases. From the diagram, we can see that the crust thickness changes gently near the Honghe fault, while it changes acutely near the Chenghai fault where the crust thickness is deeper. Paiying station (53258), near the Honghe fault, has a shallow crust thickness of 42.1km while both Xiaoyindian and Caifeng stations, which are close to the Chenghai fault and 4km away from each other, have a crust thickness of 47.9km.
Poisson's ratio is a physical quantity describing an elastic property of the media. Generally speaking, the Poisson's ratio of rock is between 0.20-0.35, varying by the composition of the rock. As is well known, the poisson's ratio of silicon is only 0.09 and Poisson's ratio diminishes when silicon oxide content increases and the increases when high-speed rail magnesia content increases. That is to say, the change of Poisson's ratio shows a trend change of mafic and felsic and its content in the crust (especially lower crust).
Within the area chosen (Fig. 5), the distribution of Poisson's ratio is significantly uneven and different between the north and the south. The Poisson's ratio increases from the north to the south, with 0.245 at Nanjingzhuang in the north and 0.277 at Mashi in the south. In this area, the distribution of Poisson's ratio is also characteristic of lateral variation. From the Honghe fault to Chenghai faut, Poisson's ratio gradually diminishes. Paiying station, near the Honghe fault, has the highest Poisson's ratio of σ=0.300, while Caifeng station has Poisson's ratio of 0.228. The distribution of Poisson's ratio is opposite to the thickness of crust; i.e., Poisson's ratio is low near the Chenghai fault zone and high near the Honghe fault zone. From the above figure, the Poisson's ratio of the triangle area of the northwest Yunnan rhombus block is between medium to high values (0.26≤σ≤0.29), so this area is very likely to have mafic in the crust, while Poisson's ratio in Paiying station is high (σ≥0.30), so the area is very likely to have an iron-magnesia component or possibly some melting (Owens T.J. et al., 1997). The crust of Poisson's ratio is low (σ≤0.26) in the northeast of Binchuan, indicating that the crust of this area is composed chiefly of felsic components.
The links between the crust thickness and Poisson's ratio might carry important information about the formation of the earth crust and its tectonic evolution process. If the component of the crust is sole, meaning that from the surface to Moho layer it consists of only one kind of rock, then we know the crust thickness will have no effect on Poisson's ratio. If the crust becomes thick by the changing of the lower crust, Poisson's ratio has a positive correlation with the crust thickness. If the crust becomes thick by the changing of the upper crust, Poisson's ratio has an inverse correlation with the crust thickness (Liu et al., 2011). In the research by Zhang Zhongjie et al.(2005a, 2005b) and Zhang Enhui et al. (2013), it is believed that the crust thickness in Baoshan, Tengchong and Simao in Western Yunnan changes because the lower crust becomes thicker, therefore, there is a positive correlation between the crust thickness and Poisson's ratio. However, Zha Xiaohui et al. (2013) believed that there is no correlation between Poisson's ratio and the crust thickness in the entire Yunnan area, except in eastern Yunnan where the increasing Poisson's ratio results from the thickening of the crust thickness. In this paper, we have drawn a diagram of the correlation between the crust thickness and Poisson's ratio (Fig. 6) in the Binchuan area, which shows that the crust thickness and Poisson's ratio has an inverse correlation (the correlation coefficient is 0.87), suggesting that the crust thickening in the Binchuan area comes mainly from the thickening of the upper crust.
Applying the teleseismic receiving function to the study of crust structure and components is a simple, rapid and efficient method. This paper uses the teleseismic wave data from 16 seismic stations in the Binchuan area, which were built by the active source project and the Himalayan array, to obtain the body wave receiving functions of each station and finally calculate the crust thickness and the Poisson's ratio beneath the 16 seismic stations. The research area is small in size, but more stations were set up, so the spatial resolution was improved greatly. This enables us to precisely describe the change of both crust thickness and Poisson's ratio in the triangle area (the Binchuan area) of the northwest Yunnan rhombus block.
(1) The average crust thickness in the Binchuan area is 45.3km, which is basically consistent to the result of the artificial seismic sounding obtained by Kan Rongju et al. (1986). In the area, the Moho layer varies acutely, showing a pattern of deep in the north and west but shallow in the north and east. There is a 5.8km difference between the deepest and the shallowest. The deep crust is located in the Chenghai fault zone where the contour is quite dense and the change is acute, and the shallowest is in Honghe fault zone where there is a gentle change in crust thickness. Generally speaking, the crust gradually thickens from the southwest to the northeast.
(2) The Poisson's ratio in this area gradually diminishes from the southwest to the northeast. The southwest has a medium-high Poisson's ratio (0.26≤σ≤0.29) while the northeast has a low Poisson's ratio (σ≤0.26). The distribution of Poisson's ratio is opposite to the crust thickness, which shows that in the area the crust thickens mainly by the upper crust.
(3) In such a small research area, we find a relatively great difference between the crust thickness and Poisson's ratio, owing to the distribution of teleseismic locations (many in the northeast and the southeast, few in the northwest and the southwest). As the interface tilts, the remote events from different directions bring out different receiving functions which cause a big difference in calculating the crust thickness and Poisson's ratio. Moreover, it is relatively difficult to distinguish the multiple waves, therefore, we take the method of mutual confirmation instead, which also causes a difference. Generally speaking, applying receiving functions to the study of the crust structure beneath the seismic station is yet a simple and mature method.
Feng Jing, Fu Zhuwu, Gao Mengtan. S-wave velocity structure in Tengchong area inversed by receiver functions[J]. Progress in Geophys., 2012, 27(4): 1328–1336. |
He Chuansong, Wang Chunyong, Wu Jianping. A study on the deep structure using teleseismic receiver function in western Yunnan[J]. Acta Seismologica Sinica, 2004b, 26(3): 238–246. |
He Chuansong, Wang Chunyong, Wu Jianping. S-wave velocity structure inferred from receiver function inversion in Tengchong volcanic area[J]. Acta Seismological Sinica, 2004a, 26(1): 11–18. |
He Xiang, Wu Guohua. Crustal thickness and its function of isostatic gravity compensation in Yunnan area[J]. Journal of Seismological Research, 1986, 9(4): 487–500. |
Hu Jiafu, Su Youjin, et al. S-wave velocity and Poisson's ratio structure of crust in Yunnan and its implication[J]. Science in China (Ser.D), 2003, 33(8): 714–722. |
Institute of Geology, S SB. Seismological Bureau of Yunnan Province[M]. Active Faults in the Western Yunnan[M]. Beijing: Seismological Press, 1990: 242-247. |
Kan Rongju, Lin Zhongyang. A preliminary study on crustal and upper mantle structures in Yunnan[J]. Earthquake Research in China, 1986, 2(4): 50–61. |
Langston C.A. Structure under Mount Rainer, Washington, inferred from teleseismic body waves[J]. J Geophys Res, 1979, 84(B9): 4749–4762. DOI:10.1029/JB084iB09p04749. |
Li Yonghua, Wu Qingju, et al. Crustal structure in the Yunnan region determined by modeling receiver functions[J]. Chinese Journal of Geophys, 2009, 52(1): 67–80. |
Liu Q.L., Wang C.Y., Yao Z.X., et al. Study on crustal thickness and velocity ratio in mid-western North China Craton[J]. Chinese Journal of Geophys., 2011, 54(9): 2213–2224. |
Liu Qiyuan, Chen Jiuhui, Li Shuncheng, et al. A passive seismic experiment in Xingjiang-Jiashi strong earthquake region and discussion on its seismic genesis[J]. Chinese Journal of Geophysics, 2000, 43(3): 356–364. |
Liu Qiyuan, Kind R., et al. Maximal likelihood estimation and nonlinear inversion of the complex receiver function spectrum Ratio[J]. Acta Geophysica Sinica, 1996, 39(4): 502–513. |
Liu Qiyuan, Li Shuncheng, et al. Broadband seismic array study of the crust and upper mantle velocity structure beneath Yanhuai basin and its neighboring region[J]. Acta Geophysica Sinica, 1997, 40(6): 763–771. |
Owens T.J., Zandt G. Implications of crustal property variations for model of Tibetan Plateau evolution[J]. Nature, 1997, 387: 7–43. |
Wu Jianping, Ming Yuehong, Wang Chunyong. The S wave velocity structure beneath digital seismic stations of Yunnan Province inferred from teleseismic receiver function modeling[J]. Chinese Journal of Geophysics, 2001, 44(2): 228–237. |
Wu Qingju, Zeng Rongsheng. The crustal structure of Qinghai-Xizang plateau inferred from broadband teleseismic waveform[J]. Acta Geophysica Sinica, 1998, 41(5): 669–679. |
Yuan X., Ni J., Kind R., et al. Lithospheric and upper mantle structure of southern Tibet forms a seismological passive source experiment[J]. Journal of Geophysical Research, 1997, 102(B12): 27491–27500. DOI:10.1029/97JB02379. |
Zandt G., Stephen C., Terry C.W. Crust and mantle structure across Basin Range-Colorado Plateau boundary at 37°N latitude and implications for Cenozoic extensional mechanism[J]. J Geophys Res, 1995, 100(B6): 10529–10548. DOI:10.1029/94JB03063. |
Zha X.H., Lei J.S. Crustal thickness and Poisson's ratio beneath the Yunnan region[J]. Scientia Sinica(Series D:Earth Sciences), 2013, 43(3): 446–456. |
Zhang E.H., Lou H., Jia S.X., et al. The deep crust structure characteristics beneath western Yunnan[J]. Chinese Joural Geophys., 2013, 56(6): 1915–1927. |
Zhang Xiaoman, Hu Jiafu, Hu Yili, et al. The S-wave velocity structure in the crust and upper mantle as well as the tectonic setting of strong earthquake beneath Yunnan region[J]. Chinese Journal of Geophysics, 2011, 54(5): 1222–1232. |
Zhang Z.J., Bai Z.M., Wang C.Y., et al. Crustal structure of Gondwana-and Yangtze-typed blocks:An example by wide-angle seismic profile from Menglian to Malong in western Yunnan[J]. Science in China (Series D:Earth Sciences), 2005a, 48(11): 1828–1836. DOI:10.1360/03yd0547. |
Zhang Z.J., Bai Z.M., Wang C.Y., et al. The crustal structure under Sanjiang and its dynamic implications:Revealed by seismic reflection/refraction profile between Zhefang and Binchuan, Yunnan[J]. Science in China (Series D:Earth Sciences), 2005b, 48(9): 1329–1336. DOI:10.1360/01YD0567. |
Zou Zuihong, Chen Xiaofei. Mapping crustal S-wave velocity structure with SV-component receiver function method[J]. Acta Seismologica Sinica, 2003, 38(2): 15–23. |