The Chifeng-Kaiyuan fault, which extends in the EW direction along 42°N, is the dividing line between two Grade Ⅰ plates of the Xingmeng orogenic belt and the North China Platform (Ma Xingyuan, 1987, 1989). The fault is located in the lithosphere thickness variation zone, and the difference in the lithosphere thickness between both sides is significant (Lu Zaoxun et al., 2005). It is also the geographical demarcation of the North China block and the Northeast China block in seismic activity analysis (Meng Xiansen et al., 2007). At 17:11 p.m. on April 22, 2013, Beijing time, the Tongliao MS5.3 earthquake occurred on the Inner Mongolia side of the middle section of the Chifeng-Kaiyuan fault. The macroscopic epicenter is located in Ganqika Town, Zuoyi Houqi, Horqin (42.98°N, 122.36°E). On January 21, 2013, an MS3.9 (ML4.4) earthquake occurred in the same location, considering the spatial location and the 3-month recurrence interval, it can be preliminarily determined as the foreshock of the magnitude 5.3 earthquake. It is worth noting that two days after the MS3.9 earthquake, an MS5.1 earthquake occurred at Dengta, Shenyang about 170km southeast of the above-mentioned location, which caused a foreshock attribution problem of the MS3.9 earthquake. According to the definition of foreshock (Lu Yuanzhong et al., 1985; Zhu Chuanzhen et al., 1996; Chen Yuntai et al., 2000; Wang Linying et al., 2005; Jiang Haikun et al., 2006), the closeness of the spatial position is more indicative than the closeness of the occurrence interval. The MS3.9 earthquake can therefore be presumed as the direct foreshock of the Tongliao MS5.3 earthquake, due to their close epicenters, although the MS3.9 earthquake and Dengta MS5.1 earthquake are only two days apart. In order to determine the nature of the foreshock more accurately, on the basis of statistical analysis of earthquake sequence, this paper describes the spatial distribution of the foreshock and the mainshock via HypoDD, examines the homology of the foreshock and the mainshock hypocenter, analyzes the focal mechanism similarity of the foreshocks through the focal mechanism analysis and statistical analysis of the seismic sequence, and comprehensively determines the relationship between the ML4.4 earthquake sequence on January 21, 2013 and the Tongliao MS5.3 earthquake.1 EARTHQUAKE OVERVIEW 1.1 Seismic Geological Background
The Songliao Basin is a typical Mesozoic-Cenozoic continental extensional, fault-depression compound basin. Its formation and evolution are mainly controlled by two kinds of dynamics: the thermal dynamics of the deep mantle material in the crust and the dynamic of the subduction of the Pacific Plate into the Asian continent. Early stage development of the basin is mainly controlled by the first type of dynamics, and the middle and late stages are mainly controlled by the second type of dynamics. Due to changes in the nature of the two dynamics, there is duality of compression and extension during the development of the basin (Chi Yuanlin et al., 2002; Liu Dianmi, 2008). At present, the Songliao Basin is still mainly affected by the westward subduction of the western Pacific plate, resulting in the formation of a medium-strong seismic activity zone in the middle and periphery of the basin. In the recorded history of earthquakes, there are 8 earthquake with MS≥5.0 having occurred within 150km of the epicenter. The earliest recorded earthquake is the MS5.0 earthquake in Fuxin, Liaoning Province in 1318, which is about 41.5km away from the Tongliao MS5.3 earthquake in 2013. The closest earthquake in time and spatial distance is a MS6.0 earthquake in Tongliao, Inner Mongolia on July 9, 1942, about 35km away from the MS5.3 earthquake in Tongliao in 2013 (Fig. 1).
According to the official earthquake catalog provided by the National Seismological Cataloging System (http://10.5.202.22/bianmu/index.jsp), from January 1, 2000 to May 14, 2013, a total of 82 earthquakes were recorded at the epicentral region of the Tongliao MS5.3 earthquake (122.25°-122.45°E, 42.88°-43.08°N). It is evident from its time series that the Tongliao MS5.3 earthquake was a medium-strong earthquake that occurred during the seismic enhancement stage within the seismically quiet area. Since the occurrence of enhancement of earthquake activity, that is, from January 21, 2013 to May 14, 2013, 79 earthquakes have been recorded in the epicentral area of the Tongliao MS5.3 earthquake, of which there are 1 earthquake with ML5.0-ML5.9, 2 earthquakes with ML4.0-ML4.9, 15 earthquakes with ML3.0-ML3.9, 43 earthquakes with ML2.0-ML2.9, and 18 earthquakes with ML1.0-ML1.9. The ML4.4 earthquake on January 21 was 73 days before the MS 5.3 earthquake on April 22, this time interval is between the"direct foreshock"and the"generalized foreshock"(Lu Yuanzhong et al., 1985; Mei Shirong et al., 1993; Wang Linying et al., 2005). In order to investigate the spatial distribution accurately, the HypoDD location method was used to relocate the seismic events in the epicentral region.2 EARTHQUAKE SEQUENCE RELOCATION
Lu Zaoxun et al. (2002) have conducted in-depth studies on the upper crust and mantle structure in the Northeast China region based on artificial seismic exploration data, and the results are relatively reliable. This paper uses the tomographic imaging results of Lu Zaoxun et al. (2002) to construct a stratified velocity structure (Table 1). Based on the results of the analysis of H-k stacking of receiver functions by Wei Zigen et al. (2012), the ratio of the wave velocity in this zone was set to 1.77.
To accurately analyze the temporal and spatial distribution of aftershock activity, HypoDD location of seismic sequences was performed using the conjugate gradient algorithm (Waldhauser F. et al., 2000; Yang Zhixian et al., 2003; Huang Yuan et al., 2008). Relocation results show that the average location errors of the EW, NS, and UD directions are 0.8km, 0.7km, and 1.1km, and the distribution ranges are 0.31km to 1.29km, 0.25km to 1.53km, and 0.34km to 2.11km, respectively. Horizontal location accuracy significantly excels vertical location accuracy. The distribution of the travel time residual is 0.262s to 0.823s. After relocating, the focal depth is concentrated at 5.5km-7.5km. It can be seen from the focal depth profile of a1-a2 that the focal depth of the sequence shows a trend of gradual deepening from northwest to southeast on the whole (Fig. 3(d)), the focal depths of the six events of the foreshock sequence ranged from 5.8km to 8.1km, and the vertical difference of the focal positions was not significant (Fig. 3(c)). After relocating, the seismic sequence showed an overall NW-trending extension of about 10km. Aftershocks were distributed on both sides of the mainshock, showing bilateral fracture characteristics; ML≥3.0 aftershocks were concentrated at approximately 2km south of the mainshock. In the foreshock sequences, the 4 events with larger magnitudes were banded in the northwest direction, and the 2 events with smaller magnitudes were scattered significantly. It's possibly due to the larger locating error of smaller earthquake events (Fig. 3(a)). The epicenters of the MS5.3 mainshock and the ML4.4 foreshock are about 1.8km apart, and the focal depths are 7.208km and 7.089km, respectively, indicating that the source locations are relatively close and the two events maybe occurred on the same fault plane.
Rock mechanics experiments and focal physics studies have shown (Mogi K., 1962; Nur A., 1972; Anderson D.L. et al., 1973; Scholz C.H. et al., 1973; Mjachkin V.I. et al., 1975) that the medium in the focal region would undergo a non-elastic deformation and exhibit an apparent anisotropy. This anisotropy may manifest in different ways, such as similar waveforms, dominant or aligned orientations of fractures within the medium, and similar focal mechanism solutions (Sobolev G.A., 1979; Cui Zijian et al., 2012). In order to further determine the foreshock sequence of the Tongliao MS5.3 earthquake, the source mechanism similarity of the foreshock sequence was investigated by solving the focal mechanism as below.
The CAP method (Cut and Paste, CAP) (Zhu et al., 1996) was used for the focal mechanism solution of MS5.3 mainshock. The P-wave and S-wave amplitude ratio methods based on the P-wave initial motion (Liang Shanghong et al., 1984; Liu Jie et al., 2004; Hu Xinliang et al., 2004; Zhang Yongjiu et al., 2007) were used for the focal mechanism solution of the foreshocks and aftershocks. These two methods have a relatively weak dependence on the station distribution (Lin Jizeng et al., 1991; Hu Xinliang et al., 2004; Long Feng et al., 2010).
The detailed calculation principle of the CAP method and the amplitude ratio method can be found in the literature. Here, the overall process of the two methods for solving the focal mechanism and related constraints are explained as follows. (1) During the CAP method solving process, the EVT waveform record of the seismic event is selected. Firstly, we remove the instrument responses recorded at stations 150km to 500km away from the epicenter, rotate the recorded waveforms to the radial, tangential and vertical directions, and then decompose them into two parts, Pnl and Snl. Suppress the noise parts of Pnl and Snl in 0.05Hz-0.2Hz and 0.05Hz-0.1Hz frequency bands respectively using the four order band-pass filter, and weight them differently. The widely used and mature method, frequency-wave number method (F-K) is used to calculate the Green's function of the epicentral distance to obtain a synthetic seismogram. Then, cross-correlate the theoretical seismogram and the actual observation waveform, and finally the focal point search method is used to invert the focal mechanism. In order to weaken the influence of factors such as velocity model and inaccurate seismic location on the results, and ensure the accuracy of inversion results, different weights are given to Pnl and Snl in the process of waveform cross-correlation and different time shifts are adopted. (2) Secondly, in the process of amplitude ratio calculation, the one-source dislocation model in layered media is used to calculate the synthetic seismogram and obtain the maximum amplitudes of P- and S-wave, and the ratio of the maximum amplitude of P- and S-wave is fitted to the observed data, in order to determine the focal mechanism. To ensure the stability and reliability of the focal mechanism solution, more than 5 stations participated in the calculation, whose accuracy is controlled within 0.2mm. The recorded amplitude is greater than 0.5mm, and is not limited, and the epicenter distance is controlled within 150km. On this account, it can be ensured that Pn, P11, Sn, S11 and other phases are not involved in the calculation.
According to the seismic data since 1970, the maximum curvature method (Woessner J. et al., 2005) was used to fit the G-R curve of the ML1.5-5.0 earthquakes occurring within 120km of the epicenter, and a minimum integrity magnitude Mc=ML2.0 of the region was obtained (Fig. 5). Since 2007, the Inner Mongolia Digital Seismic Network has begun to incorporate seismic stations in neighboring provinces and regions in locating earthquakes. The earthquake monitoring capability in the epicentral area is currently ML1.5, and the earthquake location accuracy is Class Ⅰ.
Based on above stated principles and methods, we selected the waveform records of 15 stations with high signal-to-noise ratio and epicentral distances in the range of 150km-500km for preliminary inversion of moment tensors (Fig. 4), and selected 9 stations whose actual observed waveforms have smaller fitting error with theoretical waveforms. The tensor inversion was performed again at these 9 stations, and the focal mechanism solution for the MS5.3 Tongliao earthquake was obtained as follows: nodal plane Ⅰ: strike 221°, dip angle 85°, rake 152°; nodal plane Ⅱ: strike 314°, dip angle 62°, rake 6°, basically a pure strike-slip focal mechanism, which is basically consistent with other domestic and foreign research results (Table 2, Fig. 6).
Considering the actual distribution of seismic stations (Fig. 4), it is difficult to obtain precise focal mechanism solutions for the smaller-magnitude foreshocks and aftershocks, which may affect the subsequent analysis of the results; therefore, for ML≥3.0 foreshocks and MS≥3.0 aftershocks, P-wave initial amplitude ratio method is used uniformly for solving the focal mechanism. The results show that the three foreshocks are all normal fault types, consistency is observed from the focal mechanisms of the ML3.2 and ML3.5 earthquakes, and the parameters of the nodal planes and stress axes are relatively close (Table 2, Fig. 7). Comparatively, the focal mechanism type of the aftershocks varies greatly. The focal mechanism type of the two early aftershocks is similar to that of the mainshock, which is strike-slip. The types of the latter two aftershocks are the thrust type and the normal fault.
Regarding the similarity of the focal mechanism of the foreshocks, Chen Yong (1978) studied the characteristics of the foreshocks and aftershocks and found that the foreshocks are spatially concentrated and the focal mechanism solutions are also similar, while the aftershocks are spatially dispersed and the focal mechanism solutions are different. Chen believed that the main factor causing consistency and confusion of focal mechanism is whether or not the basic structure of the rock formations or structures changes before and after the mainshock. When the external stress remains stable, geometric shape is the most powerful factor to affect the type of focal mechanism, among all the basic structures of the seismogenic structure. If the distribution of fractures in the medium is directional or shows other anisotropies, the regular geometric shape can easily produce similar focal mechanisms.
(1) The MS5.3 earthquake in Tongliao has a certain degree of particularity. It does not show the traits of moderate-strong earthquakes in the northern part of the Chifeng-Kaiyuan fault as concluded by previous research, including sudden burst and monotonic sequences (Meng Xiansen et al., 2007). Typical foreshocks occurred before the MS5.3 earthquake, and the aftershock activities were developed with a certain scale. The intensity and duration of the aftershocks exceed the historical statistics.
(2) According to the results of HypoDD relocation, the Tongliao MS5.3 earthquake sequence is distributed in the northwest direction. The focal depth shows a gradual deepening trend from the northwest to the southeast. The sequence extends approximately 10km. The aftershocks are distributed on both sides of the mainshock, with bilateral rupture characteristics. Aftershocks with larger magnitudes (ML≥3.0 aftershocks with distributed south of the mainshock. The foreshock sequence extends approximately 1.8km, and the foreshock with larger magnitudes in the sequence are mainly distributed in a northwestern belt, southeast to the mainshock. The epicentral distance of the MS5.3 mainshock and the largest foreshock (ML4.4) is about 1.8km, and the focal depths are 7.208km and 7.089km respectively. The hypocenter locations of the two earthquakes are very close and have certain "homologous" features.
(3) The focal mechanism solutions of MS5.3 mainshock, ML≥5.3 mainshocks and MS≥3.0 and mainshocks are solved by the CAP inversion method based on waveform fitting and the amplitude ratio method based on P-wave initial motion. The focal mechanism of the MS5.3 mainshock is strike-slip. The type of the aftershock focal mechanism varies greatly, and it changes from strike-slip in the early stage to thrust and normal fault in the late stage. The type of focal mechanism of foreshocks is relatively uniform and they are all normal fault. This consistency of the focal mechanism further clarified the nature of the foreshock. At the same time, it should be noted that the P-axis orientation of the ML4.4 foreshock and the subsequent two foreshocks is not consistent, and the included angle is approximately 90°. The change of the stress axis orientation in the short term may be caused by the adjustment of the stress field in the focal area itself, and may also be related to the peripheral stress triggering. 2 days after the ML4.4 earthquake, the MS5.1 Dengta earthquake occurred at about 170km south to it. This stress triggering may also lead to the change of the stress axis orientation of the above mentioned foreshocks. It should be noted that this is only the subjective speculation of the author. To accurately reveal the reasons, it is necessary to conduct research on the role of Coulomb stress loading and unloading in the epicentral area.
(4) According to the spatial and temporal distribution of earthquake sequences and the similarity of focal mechanism, the Tongliao MS5.3 earthquake is comprehensively determined as a foreshock-mainshock-aftershock type earthquake.
The figures in this paper were drawn with the Matlab programming platform and the GMT package. The CAP inversion program was written by Prof. Zhu Lupei from the University of St. Louis, USA. The peer reviewers put forward solid and sound suggestions for the modification. The authors would like to express their sincere thanks to all of them.
|Anderson D.L., Whitocomb J.H. The Dilatancy-diffusion Model of Earthquake Prediction[C]. In: Kovach R.L., Nur A. (Editors). Proceedings of the Conference on Tectonic Problems of the San Andreas Fault System[G]. Palo Alto: Stanford University Press, 1973. 417-426.|
|Chen Yong. Consistency of focal mechanism as a new parameter in describing seismic activity[J]. Acta Geophysica Sinica, 1978, 21(2): 142–159 (in Chinese with English abstract).|
|Chen Yuntai, Wu Zhongliang, Wang Peide, Xu Lisheng, Li Hongji, Mu Qiduo. Digital Seismology[M]. Beijing: Seismological Press, 2000: 83-90. (in Chinese).|
|Chi Yuanlin, Yun Jinbiao, Meng Qi'an, Yin Jinyin, Men Guangtian. Deep Structure and Dynamics of Songliao Basin and Hydrocarbon Accumulation[M]. Beijing: Petroleum Industry Press, 2002 (in Chinese).|
|Cui Zijian, Li Zhixiong, Chen Zhangli, Zhao Cuiping, Zheng Sihua, Zhou Lianqing. A study on the new method for determining small earthquake sequence type — correlation analysis of spectral amplitude[J]. Chinese Journal of Geophysics, 2012, 55(5): 1718–1724 (in Chinese with English abstract).|
|Hu Xinliang, Diao Guiling, Ma Jin, Li Xueying, Gao Jingchun, Wang Qincai. Reliability analysis of focal mechanism solutions of micro-earthquakes determined from amplitude ratio of Pˉ and Sˉ recorded by digital seismograph[J]. Seismology and Geology, 2004, 26(2): 347–354 (in Chinese with English abstract).|
|Huang Yuan, Wu Jianping, Zhang Tianzhong, Zhang Dongning. Relocation of the M8.0 Wenchuan earthquake and its aftershock sequence[J]. Science in China (Series D:Earth Sciences), 2008, 51(12): 1703–1711 . DOI:10.1007/s11430-008-0135-z.|
|Jiang Haikun, Li Yongli, Qu Yanjun, Hua Aijun, Dai Lei, Zheng Jianchang, Hou Haifeng. Spatial distribution features of sequence types of moderate and strong earthquakes in Chinese mainland[J]. Acta Seismologica Sinica, 2006, 28(4): 389–398 (in Chinese with English abstract).|
|Liang Shanghong, Li Youming, Shu Peiyi, et al. On the determining of source parameters of small earthquakes by using amplitude ratios of and from regional network observations[J]. Acta Geophysica Sinica, 1984, 27(3): 249–257 (in Chinese with English abstract).|
|Lin Jizeng, Jiang Weiqiang, Li Youming, Liang Shanghong. Determination of source parameters of small earthquakes in the east part of Guangdong and south part of Fujian Province[J]. Acta Seismologica Sinica, 1991, 13(4): 420–429 (in Chinese with English abstract).|
|Liu Dianmi. Partial Geophysical Features of Songliao Basin and Its Peripheral Typical Basins [D]. Doctoral thesis. Changchun: Jilin University, 2008 (in Chinese with English abstract).|
|Liu Jie, Zheng Sihua, Kang Ying, Chou Yongqing. The focal mechanism determinations of moderate-small earthquakes using the first motion and amplitude ratio of P and S wave[J]. Earthquake, 2004, 24(1): 19–26 (in Chinese with English abstract).|
|Long Feng, Zhang Yongjiu, Wen Xueze, Ni Sidao, Zhang Zhiwei. Focal mechanism solutions of ML≥4.0 events in the MS6.1 Panzhihua-Huili earthquake sequence of August 30, 2008[J]. Chinese Journal of Geophysics, 2010, 53(12): 2852–2860 (in Chinese with English abstract).|
|Lu Yuanzhong, Chen Zhangli, Wang Biquan. Seismological Method for Earthquake Prediction[M]. Beijing: Seismological Press, 1985: 111-150. (in Chinese).|
|Lu Zaoxun, Jiang Xiuqin, Pan Ke, Bai Yun, Jiang Delu, Xiao Liping, Liu Jianhua, Liu Futian, Chen Hui, He Jiankun. Seismic tomography in the northeast margin area of Sino-Korean platform[J]. Chinese Journal of Geophysics, 2002, 45(3): 338–351 (in Chinese with English abstract).|
|Lu Zaoxun, Jiang Delu, Bai Yun, Niu Xue, Meng Buzai. Exploration and research on the structure of the crust and upper mantle in Northeast China[J]. Seismological Research of Northeast China, 2005, 21(1): 1–8 (in Chinese with English abstract).|
|Ma Xingyuan. An Outline of the Lithospheric Dynamics of China[M]. Beijing: Geological Publishing House, 1987 (in Chinese).|
|Ma Xingyuan. Lithospheric Dynamics Atlas of China[M]. Beijing: China Cartographic Publishing House, 1989 (in Chinese).|
|Mei Shirong, Feng Deyi. Introduction to Earthquake Prediction in China[M]. Beijing: Seismological Press, 1993: 69-73. (in Chinese).|
|Meng Xiansen, Guan Yuhui, Jiang Jinhua. Difference of the earthquake sequences activity on the both sides of Kaiyuan-Chifeng fault[J]. Seismological Research of Northeast China, 2007, 23(2): 22–29 (in Chinese with English abstract).|
|Mjachkin V.I., Brace W.F., Sobolev G.A., Dieterich J.H. Two models for earthquake forerunners[J]. Pure and Applied Geophysics, 1975, 113(1): 169–181 . DOI:10.1007/BF01592908.|
|Mogi K. Study of elastic shocks caused by the fracture of heterogeneous materials and its relations to earthquake phenomena[J]. Bulletin of the Earthquake Research Institute, University of Tokyo, 1962, 40(1): 125–173 .|
|Nur A. Dilatancy, pore fluids and premonitory variations of tS/tP travel times[J]. Bulletin of the Seismological Society of America, 1972, 62(5): 1217–1222 .|
|Scholz C.H., Sykes L.R., Aggarwal Y.P. Earthquake prediction: a physical basis[J]. Science, 1973, 181(4102): 803–810 . DOI:10.1126/science.181.4102.803.|
|Sobolev G.A. Physical Processes of Earthquake Preparation: Experiment and Theory [C]. In: Proceedings of International Symposium on Earthquake Prediction [G]. Paris: UNESSCO, 1979. 281-310.|
|Waldhauser F., Ellsworth W.L. A double-difference earthquake location algorithm: method and application to the Northern Hayward fault, California[J]. Bulletin of the Seismological Society of America, 2000, 90(6): 1353–1368 .|
|Wang Linying, Chen Peiyan, Wu Zhongliang, Bai Tongxia. Characteristics of foreshock and its identification[J]. Acta Seismologica Sinica, 2005, 27(2): 171–177 (in Chinese with English abstract).|
|Wei Zigen, Chen Ling. Regional differences in crustal structure beneath northeastern China and northern North China Craton: constraints from crustal thickness and VP/VS ratio[J]. Chinese Journal of Geophysics, 2012, 55(11): 3601–3614 (in Chinese with English abstract).|
|Woessner J., Wiemer S. Assessing the quality of earthquake catalogues: estimating the magnitude of completeness and its uncertainty[J]. Bulletin of the Seismological Society of America, 2005, 95(2): 684–698 . DOI:10.1785/0120040007.|
|Yang Zhixian, Chen Yuntai, Zheng Yuejun, Yü Xiangwei. Accurate relocation of earthquakes in central-western China using the double-difference earthquake location algorithm[J]. Science in China (Series D: Earth Sciences), 2003, 46(S2): 181–188 .|
|Zhang Yongjiu, Cheng Wanzheng. Feasibility study on determining focal mechanism of small earthquakes using amplitude ratio of P- and S-waves[J]. Earthquake Research in China, 2007, 23(4): 366–374 (in Chinese with English abstract).|
|Zhu Chuanzhen, Wang Linying. Study on the characteristics of foreshock and its identification[J]. Acta Geophysica Sinica, 1996, 39(1): 80–88 (in Chinese with English abstract).|
|Zhu Lupei, Helmberger D.V. Advancement in source estimation techniques using broadband regional seismograms[J]. Bulletin of the Seismological Society of America, 1996, 86(5): 1634–1641 .|