Earthquake Reaearch in China  2017, Vol. 31 Issue (1): 39-50
Analysis of the Wave Velocity Ratio Anomalies in the Tianshan Region of Xinjiang
Zhang Linlin1, Gao Chaojun2
1 Earthquake Administration of Xinjiang Uygur Autonomous Region, Urumqi 830011, China;
2 Barkol Seismic Station, Earthquake Administration of Xinjiang Uygur Autonomous Region, Hami 839200, Xinjiang, China
Abstract: Based on the seismic observation report data provided by the Xinjiang Digital Seismic Network from 2009 to 2014, we calculate the wave velocity ratio and its background value for medium and small earthquakes by using the multi-station method in Tianshan, Xinjiang. This paper analyzes the variation of the wave velocity ratio disturbance value to highlight the abnormal, and also back-traces 7 moderate earthquakes at the research area. The results show that:(1) the background value of the wave velocity ratio is almost 1.70, the wave velocity ratio obviously decreases in the middle-eastern part of Tianshan and the region near the Puchang fault; (2) the wave velocity ratio disturbance value is mostly low in the epicenter before four earthquakes of M ≥ 5.0 from 2011 to 2013 in the study area; (3) before 7 moderate strong earthquakes, the earthquake events with low value of the wave velocity ratio account for over 60% of corresponding total events near the epicenters, and the low value of the wave velocity ratio is relatively obvious before moderate earthquakes.
Key words: Seismic observation report     Multi-station method     The wave velocity ratio disturbance value     Tianshan region

INTRODUCTION

In the process of crustal evolution, a series of changes take place in the physical state of crustal media, and seismic waves carrying a large amount of information can reflect earthquake sources and media variations, the velocity of which changes when traveling through crustal media. The wave velocity ratio, as dynamic characteristics for seismic wave, can be used to capture the information of anomalous changes before and after moderate-strong earthquakes, which is also an important basis for the study of physical property changes of underground media by using the wave velocity ratio. Through research on seismic the wave velocity ratio changes, its corresponding relationship with moderate-strong earthquakes is revealed. Therefore, using the wave velocity ratio changes as a monitoring method to predict future large earthquakes has become an important approach for seismic observation and forecasting research (Feng Deyi, 1981; Cai Xinghui, 2013; Li Haibing et al., 2008; Zhang Bo et al., 2013; Zhang Hongyan et al., 2010; Han Xiaoming et al., 2009; Long Haiying et al., 2011; Zhang Hui et al., 2014).

In 1956, by studying the Fukui MS 7.2 earthquake, Miyamoto Sadao found the existence of the wave velocity anomalies caused by head wave travel-time anomalies before the earthquake (Feng Deyi, 1981). Later, seismologists at home and abroad carried out a great deal of research on the wave velocity anomalies, and in the meantime, a variety of research methods for the wave velocity anomalies come into being (Kui Yonggang et al., 2014). In recent years, with the further promotion of digital seismological methods and the accumulation of digital seismic waveform data, research on the characteristics of the wave velocity ratio changes have gradually increased (Wang Linying et al., 2011; Han Xiaoming et al., 2010; Long Haiying et al., 2011a,2011b), research results such as spatial distribution and temporal variation characteristics of the wave velocity ratio before moderatestrong earthquakes in different study areas have also been achieved, providing a basis for the analysis of moderatestrong earthquakes tendency. The reliability of the existence of the wave velocity ratio anomalies before moderatestrong earthquakes is also further examined (Yue Xiaoyuan et al., 2008; Chen Junhua et al., 2007; Li Mingxiao et al., 2004).

With the establishment of the Digital Seismic Network for the tenth "Five-year Plan" in Xinjiang, the density of stations has increased and instrument observation precision has been improved, creating favorable conditions for the research of the wave velocity ratio in Xinjiang. Long Haiying et al.(2011a, 2011b) conducted extensive studies on the Usu MS5.1 earthquake and Hejing MS 5.6 earthquake in Xinjiang, and believed that earthquakes took place in an approximate elliptical area with an anomalous low value of the wave velocity ratio 9-10 months after the low value appeared. Gao Chaojun et al.(2013) analyzed the wave velocity ratio anomalies before and after the Xinyuan-Hejing MS 6.6 earthquake on June 30, 2012 by using digital seismic data for middle range of the Tianshan, and results showed that the earthquake took place at northeastern edge of the anomalous low value area.

In this paper, seismic data of MS ≥1.0 along the Tianshan seismic zone recorded by the Regional Digital Seismic Network in Xinjiang during January 1, 2009-September 30, 2014 is collected, and the actual value and background value of the wave velocity ratio for small-moderate earthquakes in the Tianshan area are calculated by multi-station method, for which the background value of the wave velocity ratio is the ratio of VP to VS when no moderate-strong earthquakes occur. In this article, the value of the wave velocity ratio solved based on data in 2009 is used as the background value. The actual value is the ratio of VP to VS when strong earthquakes occur. To highlight anomalous changes, the disturbance value of the wave velocity ratio is calculated (variation of actual value relative to background value). By analysis of spatial distribution changes of disturbance value of the wave velocity ratio before four moderate-strong earthquakes since 2011, the relationship between anomalous changes of the wave velocity ratio and moderate-strong earthquakes is obtained, and the matching between disturbance value changes of the wave velocity ratio and earthquakes in the Tianshan area, Xinjiang is analyzed and discussed at the same time.

1 SEISMOGEOLOGIC BACKGROUND IN THE TIANSHAN REGION IN XINJIANG

The Tianshan orogenic belt in Xinjiang is one of the world's major Paleozoic orogenic zones, including 2 collision zones; an older collision zone in the south (lying between the northern continental margin of the Tarim block and southern continental margin of middle Tianshan) and a new collision zone along the Tianshan fault (Xiao Long, 1999). Geological structures in this region are complex. Suffering from northward pushing of the Indian Ocean Plate and the squeezing effect of the Hindu Kush zone, the crust in the Tianshan region rises continuously, forming multiple thrust-fold belts in piedmont depressions on both sides or in the foreland basin (Gao Chaojun et al., 2013), among which, the middle range of Tianshan covers the east segment of south Tianshan and southern margin tectonic belt of middle Tianshan. Research on the wave velocity ratio specific to the middle-eastern part of Tianshan was carried out. Some scholars systematically studied the wave velocity ratio anomalies before and after some moderate-strong earthquakes in Xinjiang by use of simulated seismic wave data (Wang Haitao, 1989; Ao Xueming, 1987), and results showed that the wave velocity or the wave velocity ratio changed obviously before most moderate-strong earthquakes. In recent years, Long Haiying et al. (2011a, 2011b) and Gao Chaojun et al. (2013) analyzed features of anomalous changes of the wave velocity ratio before several moderate-strong earthquakes in the middle-eastern part of Tianshan. Before moderate-strong earthquakes, the wave velocity ratio formed low value areas in different shapes, and one year after the appearance of anomalies, moderate-strong earthquakes occurred within or near the border of anomaly areas. Since 2000, after gradual completion of the regional digital seismic network in Xinjiang, earthquake-monitoring capability has been improved significantly, but for the influence of layout environment, the networks are more evenly distributed in the middle-eastern part of Tianshan than in the western part of south Tianshan (Fig. 1), but earthquakes with MS ≥1.0 in Xinjiang were mostly concentrated in the middle-eastern part of Tianshan and western part of south Tianshan along the fault zone, for which, station density is higher near the Urumqi area in the middle-eastern part of Tianshan, therefore, more data for small earthquakes is recorded, while stations in the western part of south Tianshan are relatively sparsely distributed, with a higher level of moderate-strong earthquake activities, and there is less seismic data meeting the standard of the wave velocity ratio calculation than in the middle-eastern part of Tianshan. To learn more about kinematic characteristics of seismic waves in Tianshan area in Xinjiang, i.e., the wave velocity ratio changes, under the preconditions to meet technical requirements and calculation standard, the middle-eastern part of Tianshan and western part of south Tianshan are selected as study areas in this paper, in order to capture the anomaly information of the wave velocity ratio before moderate-strong earthquakes.

 Fig. 1 Distribution of seismic stations in the Tianshan area, Xinjiang and distribution of epicenters in study areas
2 RESEARCH METHODS AND PRINCIPLES

Based on the principle of the multi-station method to calculate the wave velocity ratio, using research data in the Tianshan area, Xinjiang, drawing of longitudinal wave arrival time TP i and arrival time difference between longitudinal wave and transverse wave TSi -TPi recorded by a group of seismic station i for an earthquake is done, fitted to lines for slope calculation, and by adding 1, the wave velocity ratio is obtained (Feng Deyi, 1981). Assuming the selected study area to be ideal homogeneous elastic media, computational formula using arrival time TPi and TSi of near-shock direct P-wave and S-wave recorded by multiple stations to calculate the wave velocity ratio γ, correlation coefficient R and standard deviation S is as follows:

 $\gamma =\frac{{{V}_{~\text{P}}}~}{{{V}_{\text{ }\!\!~\!\!\text{ S}}}~}=1+\frac{n\sum\limits_{i=1}^{n}{{}}\text{ }\Delta \text{ }{{T}_{i}}{{~}^{2}}-{{\left( \sum\limits_{i=1}^{n}{{}}\text{ }\Delta \text{ }{{T}_{i}}~ \right)}^{2}}}{n\sum\limits_{i=1}^{n}{{}}\text{ }\left( \Delta \text{ }{{T}_{i}}~\cdot {{T}_{\text{P}i}}~ \right)-\sum\limits_{i=1}^{n}{{}}{{T}_{\text{P}i}}~\cdot \sum\limits_{i=1}^{n}{{}}\text{ }\Delta \text{ }{{T}_{i}}}$ (1)

where, n denotes the number of seismic stations, Δ Ti =TSi -TPi is the arrival time difference between transverse wave and longitudinal wave recorded by the ith station, then

 $R=\frac{\sum\limits_{i=1}^{n}{{}}\left( {{T}_{\text{P}i}}~-\overline{{{T}_{~\text{P}}}} \right)~\cdot \text{ }\left( \Delta \text{ }{{T}_{i}}~-\text{ }\overline{\Delta \text{ }T} \right)}{\sum\limits_{i=1}^{n}{{}}\left( {{T}_{i}}~-\overline{{{T}_{~\text{P}}}} \right){{~}^{2}}-\sum\limits_{i=1}^{n}{{}}\text{ }{{\left( \Delta \text{ }{{T}_{i}}~-\text{ }\overline{\Delta \text{ }T} \right)}^{2}}}$ (2)

where, $\overline{T{{~}_{\text{P }\!\!~\!\!\text{ }}}}=\frac{1}{n}\sum\limits_{i=1}^{n}{{}}{{T}_{\text{P}i}}$, and $\overline{~\Delta ~{{T}_{i}}}~=\frac{1}{n}\sum\limits_{i=1}^{n}{{}}~\Delta ~{{T}_{i}}$

 $S = \sqrt {\frac{{{{\left( {1 - R} \right)}^2}\sum\limits_{i = 1}^n {} {\rm{ }}{{\left( {\Delta {\rm{ }}{T_i} - {\rm{ }}\overline {\Delta {\rm{ }}T} } \right)}^2}}}{{n - 2}}}$ (3)
 ${{\gamma }_{d}}~=\frac{({{\gamma }_{a}}~-{{\gamma }_{b}})}{{{\gamma }_{a}}}~$ (4)

In formula (4), γd is the disturbance value of the wave velocity ratio, γa actual value of the wave velocity ratio, γb background value. Concepts of actual value and background value are discussed in the section of introduction.

Combining previous research on the wave velocity ratio in the Tianshan region of Xinjiang, based on earthquake observation data from the Xinjiang Digital Seismic Network during 2009-2014, the multi-station method is adopted to calculate the multi-station average wave velocity ratio near epicenters in the Tianshan region (the number of stations is no less than 4). Considering the anomalous changes of calculation results of the wave velocity ratio, and to highlight the anomalies, we calculate the disturbance value for the wave velocity ratio (formula (4)), and in the meantime, analyze its temporal and spatial changing characteristics and by earthquake cases, tracing back to explore the relationship between anomalous changes of the wave velocity ratio and moderate-strong earthquakes.

3 DATA SELECTION

Data used in this article includes seismic observation data for 15, 543 earthquakes of MS ≥1.0 in the Tianshan region during January 1, 2009-September 30, 2014, provided by the Earthquake Administration of Xinjiang Uygur Autonomous Region. Based on the principle of the multi-station method, according to the standard that an earthquake event should be recorded by at least four seismic stations, altogether 9, 590 earthquake events are selected, and available data accounts for 61.7% of the total seismic data. In order to ensure the number of samples and data accuracy, and further follow the standard of standard error of the wave velocity ratio ≤0.05 and correlation coefficient ≥0.98, 1, 755 earthquakes are picked from 9, 590 earthquake events selected. Before calculating the disturbance value of the wave velocity ratio, the wave velocity ratio value should be calculated first (including the actual and background values), and data filtering and preprocessing is done with reference to the previous way of screening (Gao Chaojun et al., 2013). Using seismic data after pretreatment as mentioned above, we first set up range parameters for the study area. Grid partitioning is done in a study area of 0.2°×0.2°, and the background value of the wave velocity ratio for each grid is calculated. Background value is calculated mainly based on the time frame when no moderate-strong earthquakes occur. We then select seismic data since 2013 to acquire corresponding the actual wave velocity ratio value, calculate disturbance value of the wave velocity ratio at the same time, and analyze its temporal and spatial variation characteristics. Finally, combined with spatial distribution of the disturbance value of the wave velocity ratio before the four moderate-strong earthquakes during 2011-2013, we analyze change features of the disturbance value of the wave velocity ratio before moderate-strong earthquakes. Calculation of the disturbance value of the wave velocity ratio is mainly based on seismic data satisfying the criteria in the study area, which generally show negative low-value anomalies.

4 RETROSPECTIVE STUDY OF EARTHQUAKE CASES 4.1 Background Value of the Wave Velocity Ratio in the Tianshan Region of Xinjiang

Liu Wenxue et al. (2011) estimated the wave velocity ratio beneath 90 seismic stations in Xinjiang and its surrounding areas. The estimated results are in the range of 1.61-1.89, with an average of 1.76. Combining previous research results, the multi-station method is used in this paper to explore the wave velocity ratio in the Tianshan region of Xinjiang. The results show that the mean value of the wave velocity ratio is about 1.70 in the study area, and the wave velocity ratio changes within a range of 1.65-1.78.

4.2 Seven Earthquake Events in the Tianshan Region

In order to highlight anomalous changes of spatial distribution of the wave velocity ratio in the study area, the study area is divided into two sub-study areas including the middle-eastern part of Tianshan and western part of south Tianshan. Fig. 2 (a) and Fig. 2 (b) display spatial distribution of the disturbance value of the wave velocity ratio in the middle-eastern part of Tianshan and western part of south Tianshan during January 2013-September 2014. The results reveal that disturbance value of the wave velocity ratio shows low-value anomalies in the Wenquan-Xinyuan area and south Urumqi area in the middle-eastern part of Tianshan and in Kalpin block in western part of south Tianshan, for which the low disturbance value of the wave velocity ratio in the west Xinyuan area forms an irregular oval, while the low disturbance value near Urumqi and in the west of Puchang fault in the western part of south Tianshan changes greatly, and its radiation scope is relatively concentrated.

 Fig. 2 Distribution map of disturbance value of the wave velocity ratio in Tianshan region during January 2013-September 2014 (a) Middle-eastern part of Tianshan; (b) Western part of south Tianshan)

By calculation, time-sequence diagram for the wave velocity ratio in the above two regions is obtained (Fig. 3), for which, the two areas framed by the dotted line in Fig. 2 (a) correspond with the contents of the time-sequence diagram respectively in Fig. 3 (a) and Fig. 3 (b). Combined with research results of three earthquake events in study areas, the wave velocity ratio presents some anomalous variation characteristics before moderate-strong earthquakes, among which, the wave velocity ratio for two earthquakes in middle-eastern part of Tianshan overall shows a decline-low value-recovery-earthquake occurrence trend, which lasts for about three months, with anomalous change amplitude of about 5%. The anomalous increase is especially relatively large, which happens in the process of low value-recovery-earthquake occurrence. The low value-recovery process appears before most earthquakes: for example, a wave velocity ratio low value anomaly appeared before the Changji MS 5.6 earthquake on March 29, 2013 and the Urumqi MS 5.1 earthquake on August 30 in southern Urumqi region (Fig. 3 (a), Fig. 3 (b)), and earthquakes took place during the recovery process of the wave velocity ratio high value. The wave velocity ratio in Kalpin block remained in a state of low value before the Kalpin MS 5.3 earthquake on December 1, 2013 (Fig. 3(c)), which presented state of low value-recovery process later in time-sequence diagram of 3 sub-areas, while no earthquake occurred, indicating that the corresponding ratio of the anomaly by using this method is not 100%. Presently, the time-sequence curve for the wave velocity ratio in southern Urumqi region, Wenquan-Xinyuan region and Kalpin block shows a variation trend of decline-low value-recovery, which may be a manifestation before the occurrence of moderate-strong earthquakes and also indicates that underground media in these three subareas may be in the process of stress accumulation, requiring further attention.

 Fig. 3 Time-sequence diagram of the wave velocity ratio in 2 anomalous disturbance value areas during January 2013-September 2014 (a) Wenquan-Xinyuan region; (b) Southern Urumqi region; (c) The Kalpin block

With the increase of density of the digital seismic monitoring network, local monitoring ability has been gradually improved, and the minimum magnitude of completeness of earthquakes that can be monitored in study area has decreased, especially in the middle-eastern part of Tianshan where stations are more evenly distributed. Since 2011, altogether 16 earthquakes with MS ≥5.0 have taken place in the Tianshan area of Xinjiang, where moderate-strong earthquakes frequently occur. In this article, variation of disturbance value of the wave velocity ratio before four moderate-strong earthquakes with MS ≥5.0 during 2011-2013 is analyzed, located in the middle-eastern part of Tianshan and western part of south Tianshan. It can be seen from the spatial distribution map (Figs.4, 5, 6) of disturbance value of the wave velocity ratio in middle-eastern part of Tianshan that the Nilka-Gongliu MS 6.0 earthquake on November 1, 2011 occurred in an area with an anomalous low disturbance value (Fig. 4). From January 2009 to November 1, 2011 before the MS 6.0 earthquake, near the epicenter and its surrounding area there were altogether 894 earthquakes that satisfied conditions for the wave velocity ratio calculation. And from January 2011, altogether 267 earthquakes had a wave velocity ratio lower than 1.71, accounting for 65% of the total number of small-moderate earthquake events before the MS 6.0 earthquake. Before the Xinyuan-Hejing MS 6.6 earthquake on June 30, 2012, the disturbance value of the wave velocity ratio formed a low value zone in an irregular semi-elliptical shape near the Xinyuan region, and the epicenter of the main shock was located at the end of the semi-ellipse (Fig. 5). Before the MS 6.6 earthquake, from January 2009 to June 30, 2012, there were altogether 1, 055 earthquakes that satisfied conditions for wave velocity ratio calculation in corresponding study area. After the MS 6.0 earthquake on November 1, 2011, altogether 106 earthquakes had a wave velocity ratio lower than 1.71, accounting for 63.5% of the total number of small-moderate earthquake events before the main shock. The anomalous low value of the wave velocity ratio before the MS 6.0 earthquake on November 1, 2011 in this area may still be related to the occurrence of the MS 6.6 earthquake. Spatial variation of the wave velocity ratio disturbance value in the middle-eastern part of Tianshan from January 2009 to March 2013 (Fig. 6) shows that the Urumqi-Changji MS 5.6 earthquake on March 29, 2013 took place in an area of concentrated low disturbance value, and from January 2009 until this earthquake, there were altogether 1318 earthquakes that satisfied conditions for the wave velocity ratio calculation near the epicenter and its surrounding areas. From January 2013 until the occurrence of the MS 5.6 earthquake, a total of 23 earthquake events had a wave velocity ratio lower than 1.72, accounting for 62% of the total number of small-moderate earthquake events before the MS 5.6 earthquake. Fig. 7 provides the spatial variation of the wave velocity ratio disturbance value in the southern Akqi region in the western part of south Tianshan from January 2009 to July 2011. The Atushi-Jiashi MS 5.6 earthquake on August 11, 2011 occurred between the two areas with low disturbance value, and from January 2009 until the occurrence of this earthquake, there were altogether 665 earthquake events near the epicenter and its surrounding area that can be used to calculate the wave velocity ratio. However, from January 2011 until the occurrence of this MS 5.6 earthquake, altogether 88 earthquake events had a wave velocity ratio lower than 1.73, accounting for 52.4% of the total number of small-moderate earthquake events before the MS 5.6 earthquake. It can be seen from the diagram of variation of disturbance value of the wave velocity ratio before these four moderate-strong earthquakes that the wave velocity ratio anomalies were mostly distributed on both sides of fault zone or intersection part between faults in different directions. Most faults have an earthquake controlling background, with underground media of very complicated physical properties, and the stress accumulation level is relatively high in area with anomalous low wave velocity ratio value and its surrounding area, which supplies good geological-tectonic conditions for the occurrence of moderate-strong earthquakes.

 Fig. 4 Distribution map of disturbance value of the wave velocity ratio near Xinyuan region from January 2009-October 2011

 Fig. 5 Distribution map of disturbance value of the wave velocity ratio near Xinyuan region from January 2009-June 2012

 Fig. 6 Distribution map of disturbance value of the wave velocity ratio near Urumqi region from January 2009-March 2013

 Fig. 7 Distribution map of disturbance value of the wave velocity ratio in southern Akqi region from January 2009-July 2011
5 CONCLUSION AND DISCUSSION

In this article, by calculating the seismic wave velocity ratio in the Tianshan region of Xinjiang since 2009, spatial and temporal variation diagram of the wave velocity ratio in the Tianshan region is achieved, and in combination with four earthquake events during 2011-2013, we analyze some changes of disturbance value of the wave velocity ratio before moderate-strong earthquakes, and the following conclusions are achieved:

(1) The average value of the wave velocity ratio in the Tianshan region of Xinjiang during 2009-2014 is 1.70, fluctuating within the range of 1.65-1.78. In combination with analysis of three earthquake events, the wave velocity ratio presents certain anomaly characteristics before moderate-strong earthquakes, with variation amplitude about 5%. It shows the changing process of decline-low value-recovery-earthquake occurrence, which lasted for about three months. Comprehensive analysis of the wave velocity ratio changes before seven moderate-strong earthquakes is done. Results show that the number of earthquake events with a low wave velocity ratio value accounts for more than 60% of the total number of corresponding earthquake events at the epicenter and its surrounding areas before 7 earthquakes, and the wave velocity ratio shows an obvious low value before moderate-strong earthquakes.

(2) According to temporal and spatial variation of disturbance value of the wave velocity ratio in the study area since 2013 and analysis of four earthquake events, an anomalous low disturbance value of the wave velocity ratio appeared in the Wenquan-Xinyuan region and southern Urumqi region in the middle-eastern part of Tianshan and Kalpin block in western part of south Tianshan. In these two subareas, seismicity level is high, with complex underground structures, and high and low disturbance values of the wave velocity ratio appeared alternately. The size of the distribution area and degree of concentration are related to station density and earthquake events that satisfy conditions for the wave velocity ratio calculation. In regions that have different seismicity levels, its underground media also shows differences in stress state changes and functions. In regions that have greater disturbance value changes, its geological tectonic stress changes are correspondingly significant, and the stress accumulation is in a process of gradually strengthening, therefore, there is a greater probability for moderate-strong earthquakes, thus further tracing and studying on changes of the wave velocity ratio and other seismic parameters in these subareas should be strengthened. In the meantime, in combination with analysis of temporal and spatial variation of disturbance value of the wave velocity ratio for four earthquake events, most moderate-strong earthquakes took place in areas with concentrated low disturbance values. Disturbance value anomalies of the wave velocity ratio were mostly distributed on both sides of fault zone or intersection part between faults in different directions, while stress changes and interactions in underground media are significant in corresponding study areas, which supplies good geological-tectonic conditions for the occurrence of moderate-strong earthquakes.

(3) Because the research data used in this article begins from 2009, seismic data from previous simulations is not adopted. Therefore, changes of the wave velocity ratio in the Tianshan region of Xinjiang cannot be presented completely in recorded history. In addition, compared with seismic stations in middle-eastern part of Tianshan, stations in the western part of south Tianshan are few and scattered, unevenly distributed, region partition precision is not sufficiently high, and earthquake events recorded are also relatively fewer, thus obtained results are greatly influenced by data. With the increasing quantity of digital seismic data and the improvement of data quality, the accuracy and reliability of research results on the wave velocity ratio will be further improved.

ACKNOWLEDGEMENTS

Research data is provided by the Earthquake Administration of Xinjiang Uygur Autonomous Region. We'd like to show our gratitude to Li Zhihai and Wang Qiong for their suggestions and advice on the calculation of the disturbance value of the wave velocity ratio and earthquake example analysis respectively. We also benefitted greatly from opinions and suggestions proposed by other reviewers.

REFERENCES
 Ao Xueming, Wang Guiling. The variations of the velocity ratio of seismic waves before and after some moderate strong earthquakes in Xinjiang[J]. Inland Earthquake, 1987, 1(1): 67–76. Cai Xinghui. The measurement and research on the average wave velocity ratio in Fujian Province[J]. Plateau Earthquake Research, 2013, 25(4): 31–34. Chen Junhua, Gan Jiasi, Li Puli, et al. On the wave velocity ratio (VP/VS) in Badong Area after impoundment of Three Gorges Reservoir[J]. Journal of Geodesy and Geodynamics, 2007, 27(Special): 12–14. Feng Deyi. The Anomalies of Seismic Velocity[M]. Beijing: Seismological Press, 1981: 1-14. Gao Chaojun, Song Chunyan, Zhang Linlin, et al. Study on characteristics of velocity ratio anomaly before and after Xinyuan Hejing Earthquake with MS6.6[J]. Inland Earthquake, 2013, 27(2): 121–126. Han Xiaoming, Wang Xin, Hao Meixian. Analysis on seismic activity and velocity ratio (VP/VS) in Chifeng and its adjacent regions[J]. Plateau Earthquake Research, 2010, 21(3): 11–16. Han Xiaoming, Zhang Jianzhong, Xue Ding. Study of the average velocity ratio (VP/VS) in the Hohhot-Baotou area[J]. Earthquake Research in China, 2010, 26(4): 418–425. Li Haibing, Wang Zongxiu, Fu Xiaofang, et al. The surface rupture zone distribution of the Wenchuan earthquake(MS8.0) happened on May 12th, 2008[J]. Chinese Geology, 2008, 35(5): 803–813. Li Mingxiao, Zhang Xiaodong. Determining average seismic velocity ratios (VP/VS) in the curst in North China Region by multi-station method[J]. Earthquake, 2004, 24(1): 163–169. Liu Wenxue, Liu Guizhong, Zhou Gang, et al. Crustal thickness and VP/VS ratio variations of Xinjiang and surrounding regions constrained by receiver function stacking[J]. Chinese Journal of Geophysics, 2011, 54(8): 2034–2041. Long Haiying, Nie Xiaohong, Tang Lanlan. Case analysis of wave velocities ratio abnormity before Hejing M5.6 earthquake in Xinjiang[J]. Earthquake Research in China, 2011a, 27(2): 147–154. Long Haiying, Nie Xiaohong, Tang Lanlan. Study of the abnormity of wave-velocity ratio before MS5.1 Wusu earthquake in Xinjiang[J]. Journal of Seismological Research, 2011b, 34(2): 126–130. Wang Haitao. Seismic velocity variations of single station[J]. Inland Earthquake, 1989, 3(2): 181–183. Wang Linying, Li Yan'e, Li Fang, et al. Temporal variation of VP/VS, VP and VS before and after 2008 Wenchuan earthquake[J]. Acta Seismologica Sinica, 2011, 33(1): 1–14. Wei Yonggang, Chen Xuezhong, Guo Xiangyun, et al. Seismic velocity variations of single station before the 2013 Lushan 7.0 earthquake[J]. Earthquake, 2014, 34(2): 105–114. Xiao Long. A review of structure evolution of Tianshan orogen, Xinjiang, China[J]. Journal of Guilin Institute of Technology, 1999, 19(4): 315–320. Yue Xiaoyuan, Wang Linying, Wu Anxu. Study on the relationship between abnormal charge of the wave velocity ratio in Beijing area and medium earthquakes in Capital zone[J]. Seismological and Geomagnetic Observation and Research, 2008, 29(3): 10–13. Zhang Bo, Feng Jiangang, Zang Hui, et al. Study on the wave velocity ratio(VP/VS)variation characteristics of Minxian-Zhangxian MS6.6 earthquake sequence, Gansu Province[J]. Northwestern Seismological Journal, 2013, 35(3): 557–561. Zhang Hongyan, Gao Jinzhe, Lu Yanhong. The measurement and research of average wave velocity ratio in Jilin Province[J]. Seismological Research of Northeast China, 2010, 26(4): 30–31. Zhang Hui, Gao Lixin, Han Xiaoming. The variation of the wave velocity ratio at the eastern region of Inner Mongolia[J]. Northwestern Seismological Journal, 2013, 35(4): 921–927.

1 新疆维吾尔自治区地震局, 乌鲁木齐市新市区科学二街338号 830011;
2 新疆巴里坤地震台, 哈密 839200