2. Institute of Geophysics, CEA, Beijing 100081, China
The Pishan MS6.5 earthquake occurred at 09:07 a.m. on July 3, 2015 in Pishan County, Hotan Prefecture, the Xinjiang Uygur Autonomous Region. Strong ground motion could be felt in the southern part of the Xinjiang region, such as the Khotan, Kashgar and Aksu areas. Three people were killed, and more than 12, 000 houses collapsed or were severely destroyed. According to China Earthquake Networks Center 2, the epicenter of the earthquake was 37.6°N, 78.2°E with a focal depth of 10km. The earthquake occurred near the Zepu fault, a buried fault with northwest striking, which is an early and middle Pleistocene fault3. The earthquake intensity of the meizoseismal area is Ⅷ with the major axis of the isoseismic line in the NWW direction 4. Harvard 5 gave the focal mechanism solution of the Pishan MS6.5 earthquake as strike, dip, and rake 109°/22°/ 85°, respectively, and the auxiliary nodal plane solution 294°/68°/92°. This mainly demonstrates the thrusting feature of the Pishan MS6.5 earthquake. The satellite image demonstrates that the earthquake was located at the front scarps of the alluvial fan, which shows a clear linear feature and reflects the traces of the fracture zone (Li Jianhua et al., 2002).
The Pishan MS6.5 earthquake was the first earthquake above MS6.0 in China after the 2015 Nepal MS8.1 earthquake. In recent years, the strong earthquakes in the vicinity of Pishan area are active, reflecting the enhanced northward compression of the Indian plate on the Eurasian plate. The Pishan area, Xinjiang is located in a low monitoring capacity area in the Chinese mainland, where the amount of research material is relatively smaller. With the completion of the digital networks, the monitoring capacity in this region has been greatly improved (Yin Guanghua et al., 2010). It can provide a reference for seismic hazard assessment through the study of the seismicity and the relocation of the Pishan earthquake sequence. Based on the seismic catalog and seismic phase report, the seismicity characteristics of strong earthquakes in the Pishan area are analyzed and the relocation of the 2015 Pishan MS6.5 earthquakes aftershocks is completed.1 TECTONIC SETTING AND SEISMICITY
The Pishan earthquake is located on the western segment of the West Kunlun Mountains and at the intersection of the Tarim Basin and the West Kunlun Mountains. As it was pushing strongly northward on the Indian plate and encountering the rigid stop of the Tarim Basin, it developed a series of dextral strike-slip faults with northwest direction in the West Kunlun area. This area belongs to the Pamir-West Kunlun tectonic belt. The western section of it spreads around the northern margin of Pamir plateau, the eastern section is connected with the East Kunlun-Altyn Tagh seismic belt along the West Kunlun, the southern area is the Qinghai-Tibetan Plateau in an intense uplift, and the northern part is the south Tianshan seismic belt with strong tectonic activity and the Kashgar-Hotan depression with strong subsidence (Hu Fangqiu, 1988). There is a huge south-dip nappe along the faults on the northern margin of the West Kunlun. It formed a piedmont depression and the thick molasse accumulation was deposited on the southern margin of the Tarim Basin as the pre-Mesozoic strata was pushed over northward above the Cenozoic strata (Wang Suyun et al., 1992).
The Western Kunlun block includes the North Kunlun tectonic belt, the middle Kunlun tectonic belt and the Karakorum tectonic belt (Cheng Jia et al., 2014). Influenced by the northward push from the Indian plate to the Eurasia plate, the tectonic activity has been more obvious in the Karakorum tectonic belt since 1890. The Taxkorgan MS7.0 earthquake on July 5, 1895 and the Hotan MS7.1 earthquake on November 19, 1996 occurred in the vicinity of the Karakorum fault, as shown in Fig. 1. Compared with the Karakorum tectonic belt, the seismicity of the MS≥7.0 earthquakes in the North Kunlun tectonic belt was not strong where the 2015 Pishan earthquake occurred. However, the moderate-strong earthquakes were active along this belt, such as the March 5, 1956 Yecheng MS5.8 earthquake, the October 27, 1975 Pishan MS5.2 earthquake, the December 24, 1995 Pishan MS5.3 earthquake, the May 31, 1997 Pishan MS5.3 earthquake, the May 29, 1998 Moyu MS6.2 and MS5.5 earthquakes, and the August 26, 2005 Moyu MS5.2 earthquake, as shown in Fig. 1.
With the construction of the Xinjiang Digital Seismic Network since 2008, the monitoring capacity of the Xinjiang area has been greatly improved. Since the 2015 Pishan earthquake, the Xinjiang Regional Digital Seismic Network has recorded a large number of aftershocks. Before 24:00 on July 31, 2015, it had recorded a total of 2, 143 earthquakes, including 1, 222 earthquakes with ML1.0-1.9, 724 earthquakes with ML2.0-2.9, 112 earthquakes with ML3.0-3.9, 28 earthquakes with ML4.0-4.9. The maximum aftershock was the ML4.8 earthquake which occurred at 09:13a.m. on July 3, as shown in Fig. 2(a).
In seismology, the b-value is often used to estimate the seismic hazard and the frequency attenuation coefficient called h-value represents how fast the earthquake sequence decays (Yang Wen et al., 2015). The h-value of this sequence was calculated at about 1.07, which indicates that the aftershock sequence decayed slowly, as shown in Fig. 2(b). According to the G-R relationship (Magnitude-Frequency), the earthquake record with above ML1.8 events was relatively complete in this Pishan earthquake sequence. From the G-R relationship of several strong earthquake sequences, the minimum magnitude of completeness is between ML2.8 and ML3.0 from 1995 to 2005, as shown in Fig. 3(b), (c), (d). The magnitude and the frequency of the earthquake sequence fit a linear relationship between ML1.8 to ML3.2 with b-value of 0.76, as shown in Fig. 3(a). Similarly, according to the linear relationship between the magnitude and frequency, the b values of the December 24, 1995 Yecheng-Pishan MS5.3 earthquake sequence, the May 29, 1998 Moyu-Pishan MS6.2, MS5.5 double earthquakes sequence, and the August 26, 2005 Moyu MS5.2 earthquake sequence are between 0.7 and 0.8. The b-value of the May 31, 1997 Pishan MS5.3 earthquake sequence has not been calculated due to the lower number of earthquakes. From the comparison of several strong earthquakes in the area before 2015, it can be seen that the monitoring capacity of the Pishan area has been greatly improved in 2015, and the b-value has not changed greatly since 1995, which fluctuated from 0.7 to 0.8.
The directivity of the earthquake sequence in the spatial distribution is not clear, as shown in Fig. 4. In recent years, especially since 1995, the tectonic activity in this area has been more intense. There were the December 24, 1995 Yecheng-Pishan MS5.3 earthquake, the May 31, 1997 Pishan MS5.3 earthquake, the May 29, 1998 Moyu-Pishan MS6.2, MS5.5 double earthquakes, and the August 26, 2005 Moyu MS5.2 earthquake near the Pishan area. In spatial distribution, these earthquakes occurred conjointly at the junction of the West Kunlun block and the Tarim basin. The directivity of the aftershocks was not clear, but it was seen that the aftershocks spread generally along the northwest direction. Harvard gave the focal mechanism solutions of several major earthquakes in the vicinity of Pishan in the Xinjiang region respectively, such as the May 31, 1997 Pishan MS5.3 earthquake, the May 29, 1998 Moyu-Pishan MS6.2 earthquake and the August 26, 2005 Moyu MS5.2 earthquake, as shown in Fig. 4. The solution demonstrates that these earthquakes are thrust-type with low-angle dip of about 22°. These reflect the dynamical features of the Tarim block plunging southward to the West Kunlun Mountains for the West Kunlun-Pamir area obducting northward on the Tarim block along the western Kunlun-Pamir northern fault. This Pishan earthquake is a typical event of low-angle subduction. This area is formed by the crustal superficial matter in the West Kunlun, which obduct on the foreland basin of Tarim for the delamination of the Qinghai-Tibetan plate (Wang Youxue et al., 2006).
In this study, we used the double-difference method (Waldhauser et al., 2000) to relocate the aftershocks accurately. This method can reduce the influence of the lateral inhomogeneity and the uncertainty of the velocity model on the position results. The domestic scholars used the double-difference method to relocate the earthquake sequences in the Xinjiang region (Huang Yuan et al., 2006; Fang Lihua et al., 2015). We relocated the Pishan earthquake sequence using the seismic phase report from July 3, 2015 to July 31, 2015 provided by the China Earthquake Networks Center. In this study, we selected the Pg and Sg phase report recorded by 22 stations including the Hotan seismic array. The distance between the stations and the aftershocks are within 300km, as shown in Fig. 4. In the double-difference method, the depth information has a great influence on the result. We selected the arrival times including 831 ML≥2.0 earthquakes with seismic depth information, which fit a clear linear relationship for the travel time of the 18451 Pg waves or the 17498 Sg waves and their distance to the stations. The average velocity of the Pg waves is 5.947km/s and the average velocity of the Sg waves is 3.472km/s by linear least squares fitting, as shown in Fig. 5. In general, because the P-wave phase is the first arrival, its reading accuracy is higher than that of the S-wave. The P-wave was given a weight of 1.0 and the S-wave a weight of 0.5. The same weights were given to the Hotan seismic array. It was about 1.71 for the average velocity ratio of the Pg-wave and the Sg-wave. This value was equivalent to the estimated velocity ratio from the H-κ superposition algorithm of the region in the receiver functions (Liu Wenxue et al., 2011). Thus, it is taken as the velocity ratio in this paper. The results of the manpower deep sounding and the receiver functions migration analysis (Li Qiusheng et al., 2000; Liu Qiyuan et al., 2000) were referred to in building the velocity model.
After relocation, a total of 713 earthquake sites were obtained. The RMS residual of the sequence was 0.02s, the error of the position was 0.029km in the EW direction, 0.020km in the NS direction, and 0.015km in the vertical direction. The aftershocks were distributed linearly along the northwest direction with a strike of 106°, which is consistent with the results of the focal mechanism solution. The main shock was located at the southeastern end of the aftershock zone, as shown in Fig. 6(a). It shows that the depth of the aftershocks was mainly 5-10km.The distribution of the aftershocks along the northwest section A-A was significantly more scattered than the northeast section B-B along the depth profile, as shown in Fig. 6(b) and Fig. 6(c). It indicates that the direction of the causative fault was northwest direction. It also can be seen that the inclination of the fault was roughly in the SW direction with a low angle from the section B-B'. The major axis of the aftershocks zone was about 50km and the minor axis of it was about 25km. The aftershocks were mainly concentrated near the main shock in the southeast. The aftershocks were scattered in the beginning and they were mainly near the Tarim Basin (B') 15 days after the main shock occurred.
It can be seen from Fig. 7(a) that the aftershock activity in the southeastern section decayed slowly and was still not complete on July 31. The aftershocks in northwest were less active and decayed quickly, and there were few aftershocks after July 28. It also can be seen from Fig. 7(b) that the aftershocks decayed with time and were accompanied by strong aftershocks along the section B-B'. The aftershock activity decreased with the distance to the main shock. The aftershocks were mainly at the B side in the beginning two days after the main shock. Aftershock activity at the B side was stronger than that at the B side two days after the main shock, which may indicate that the stress in the source zone has changed. The aftershock was distributed in a large depth range at the beginning, and then was concentrated at a depth of 6km-10km near the depth of the main shock, as shown in Fig. 7(a) and Fig. 7(b).
The 2015 Pishan MS6.5 earthquake was located at the intersection of the Tarim block and West Kunlun. In recent years, moderate-strong earthquakes were active in this area. In this paper, we studied the seismicity, seismology parameters and relocation of the seismic sequence using the earthquake catalog and the seismic phase report.
Through the calculation of seismology parameters, the b-value of this Pishan earthquake sequence is about 0.76, and the b-value of the Pishan region has not changed much since 1995. The h-value of this Pishan sequence is about 1.07, indicating that this earthquake sequence was a main shock-aftershock type event and decayed slowly. According to the focal mechanism solution and the relocation results, the Pishan earthquake was a thrust-type event along the edge of the Tarim Basin, reflecting the dynamic characteristics that the West Kunlun-Pamir area obducted northward on the Tarim block along the West Kunlun-Pamir northern fault. This area was formed by the superficial matter in the West Kunlun tectonic belt obducting on the Tarim foreland. There is a possibility of moderate-strong earthquake occurrence. After the relocation of the aftershocks, we understand that the Pishan MS6.5 earthquake is a unilateral rupture event with a triggering seismic fault, i.e. the Zepu fault with north-west strike. The major axis of the aftershock zone is about 50km long and the minor axis is about 25km long. The aftershock activity in the southeastern section was more intense and decayed slowly while that in the northwest section was less intense and decayed quickly. The aftershocks were mainly distributed close to West Kunlun a day or two days after the main shock and close to the Tarim Basin two days later. The aftershock zone gradually shrank to the main shock with time, reflecting the healing process of the fault.
ACKNOWLEGEMENT: The seismic phase report was provided by the China Earthquake Networks Center for this study. We are also grateful to Dr.Waldhauser for the hypoDD relocation program and to the reviewers for their suggestions.
This paper has been published in Chinese in the journal of Earthquake, Volume 37, Number 1, 2017.
|Cheng Jia, Liu Jie, Sheng Shuzhong, Yao Qi, Liu Daiqin. Tectonic background of the Yutian MS7.3 earthquake and its relationship with the Yutian MS7.2014.3 earthquake in 2008[J]. Earthquake Research in China, 2014, 30(2): 143–150.|
|Fang Lihua, Wu Jianping, Wang Weilai, Yang Ting, Wang Changzai. Relocation of the 2014 MS7.3 earthquake sequence in Yutian, Xinjiang[J]. Chinese J. Geophys, 2015, 58(3): 802–808.|
|Hu Fangqiu. Xinjiang seismic tectonic belts and their characteristics[J]. Inland Earthquake, 1988, 3(2): 268–276.|
|Huang Yuan, Yang Jiansi, Zhang Tianzhong. Relocation of Bachu-Jiashi, Xinjiang earthquake sequence in 2003 using the double difference location algorithm[J]. Chinese J. Geophys, 2006, 49(1): 162–169.|
|Li Jianhua, Zhang Jiasheng, Shan Xinjian. Faulting in the western Kunlun-southwestern Tianshan Area[J]. Acta Geologica Sinica, 2002, 76(3): 347–355.|
|Li Qiusheng, Lu Deyuan, Gao Rui, Li Jingwei, Fan Jingyi, Xiong Xianming, Zhang ZhiLi, Liu Wen, Li Yingkang, Yan Quanren, Li Dexingying, Liu Wen, Li Yingkang, Yan Quanren, Li Dexing. Explosion seismic probing across the West Kunlun-Tarim contact zone[J]. Science in China (Series D), 2000, 30(SI): 16–21.|
|Liu Qiyuan, Chen Jiuhui, Li Shuncheng, Guo Biao. Passive seismic experiment in Xinjiang Jiashi strong earthquake region and discussion on its seismic genesis[J]. Chinese. J. Geophysics, 2000, 43(3): 356–366.|
|Liu Wenxue, Liu Guizhong, Zhou Gang, Li Xin, Zhang Huimin, Xu Henglei. Crustal thickness and VP/VS ratio variations of Xinjiang and surrounding regions constrained by receiver function stacking[J]. Chinese J. Geophys, 2011, 54(8): 2034–2041.|
|Waldhauser F, Ellsworth W. L. A double-difference earthquake location algorithm: method and application to the Northern Hayward Fault, California[J]. Bull. Seism. Soc. Am, 2000, 90: 1353–1368. DOI:10.1785/0120000006.|
|Wang Suyun, Shi Zhenliang, Huan Wenlin. The February 14, 1980 Xinjiang Yecheng earthquake, an intermediate-depth destructive earthquake[J]. Acta Seismologica Sinica, 1992, 14(2): 137–143.|
|Wang Youxue, Jiang Mei, Xiong Shengqing, Xue Dianjun, Wittlinger G, Paul A. Delamination of the lithosphere below the West Kunlun and its tectonic implication-evidence from seismic tomographic images and aeromagnetic anomalies[J]. Geology in China, 2006, 33(2): 299–308.|
|Yang Wen, Cheng J., Liu J., Zhang X.M. The Kangding earthquake of November, 2014[J]. Earthquake Science, 2015, 28(3): 197–207. DOI:10.1007/s11589-015-0123-2.|
|Yin Guanghua, Duan Tianshan, Liu Pingren, Wu Guodong. Monitoring ability and tectonic significance of digital seismic observation network in Xinjiang[J]. Inland Earthquake, 2010, 24(2): 97–106.|
2. 中国地震局地球物理研究所，北京 100081