2. Kunming Reference Seismic Station, Yunnan Earthquake Agency, Kunming 650224, China;
3. Dali Haidong Development Municipal Construction Co., Ltd., Dali 671006, Yunnan, China;
4. Yunxian Seismic Station, Yunnan Earthquake Agency, Lincang 675803, Yunnan, China
Using seismological methods to monitor the changes in the composition, structure and state of the subsurface medium over time, and to carry out 4-D seismological studies with time-varying information is an important direction for the development of modern seismology (Chen Yong et al., 2016; Wang Weitao et al., 2017). Seismic waves can penetrate the interior of the earth and convey rich media information, thus providing conditions for internal exploration (Liu Zifeng et al., 2015). Changes in the stress state of the medium with earthquake preparation and occurrence will cause changes in the seismic wave velocity. Monitoring such wave velocity changes is of great significance for earthquake prediction and aftershock trend judgment (Wang Weitao et al., 2017).
The change of wave velocity caused by the change of stress in the underground medium is very small (Wei Yunyun et al., 2016). Accurate measurement of the change of wave velocity has high requirements on the repetitive characteristics of the source and the time accuracy of the recording instrument (Wang Weitao et al., 2009). Natural repetitive sources are limited in time, spatial resolution and accuracy (Wang Baoshan et al., 2016). Therefore, the use of artificial sources to actively launch seismic waves into the ground is another means of monitoring the changes of the underground medium. This method overcomes the insufficiency of the natural sources in time and space to some extent (mainly: the natural source is unevenly distributed in time and space; it is of complexity and passive detection). The method has the characteristics of known excitation position, controllable excitation, good repeatability and high detection accuracy. The development of high-performance artificial sources and high-precision measurement system provides conditions for high-precision detection of seismic wave velocity changes (Yang Wei et al., 2010).
The research on artificial sources has experienced a long process of development. At present, the sources mainly include explosive blasting, drop hammer, electric spark, vibrator, and airguns in water. Large-capacity airguns are characterized by green environmental protection, good repeatability, high excitation energy, and high energy conversion efficiency (Chen Yong et al., 2007a, 2007b; Wang Bin et al., 2015), which can pick up higher-precision seismic phases and obtain the high-precision wave velocity changes in the region. It is of great significance in the study of the properties of subsurface media over time. Therefore, it has been used to conduct seismic sounding studies aiming at crustal structure (Okaya D. et al., 2002; Qiu Xuelin et al., 2007). In view of the many advantages of large-capacity airguns, in recent years, under the impetus of the "Underground Light Project" by Academician Chen Yong (Chen Yong et al., 2005), large-capacity airgun sources were transplanted into land reservoirs for observation and research of underground medium changes. The construction and excitation of the Binchuan airgun source in Yunnan provides a reliable source of data for continuously tracking changes in the underground medium.
Since the establishment of the Binchuan airgun source signal launcher in 2011, many scholars have carried out research on the data reliability of the airgun source (Wang Bin et al., 2016; Chen Yong et al., 2017; Li Xiaobin et al., 2016), waveform characteristics (Zhou Qingyun, 2018; Chen Jia et al., 2016), data processing methods (Liu Zifeng et al., 2015; Luan Yi et al., 2016; Zhai Qiushi et al., 2016), data quality (Zhang Yunpeng et al., 2017) and influencing factors (Zhou Qingyun, 2018), and have achieved rich results. On the basis of the maturity of data processing methods, the Binchuan airgun signal launching station is better applied to earthquake monitoring and forecasting, and analysis of the temporal and spatial relationship between the velocity variation of the underground medium and the earthquake occurrence on the basis of the waveform data, to obtain earthquake precursor information for seismic analysis. This is of great significance in earthquake forecasting.
On May 18, 2016 and March 27, 2017, Yunlong MS5.0 (26.1°N, 99.53°E) and Yangbi MS5.1 (25.89°N, 99. 80°E) occurred in the experimental area of the West Yunnan Province. The focal depths are 15km and 12km respectively. The seismic epicenter locations are about 102km and 70km away from the active source excitation source. Both earthquake epicenters are located in the coverage of the receiving stations, and are located in the continuous excitation period in the time domain. The rich waveform data can be obtained before and after the earthquake, which is the characteristic of the velocity change of the airgun source signal before and after the earthquake with MS≥5.0. In-depth study and analysis of the temporal and spatial evolution characteristics of the stress state of the subsurface medium provide a good earthquake case.
Based on the airgun source signal recorded by the station from January 2016 to June 2017, this paper uses cross-correlation detection technology to obtain the stable phase change characteristics of each station. Taking the Yunlong MS5.0 and Yangbi MS5.1 earthquakes as samples, combined with the GNSS baseline data, the characteristics of regional wave velocity changes before and after the earthquake are analyzed in depth, and the commonalities and differences between the two earthquake cases are summarized, so as to use the airgun source signal to study the characteristics of wave velocity variations before and after the MS≥5.0 earthquakes.1 DATA COLLECTION AND PROCESSING ANALYSIS 1.1 Data Collection
The excitation water body of the large-capacity airgun source in Binchuan is the Dayindian Reservoir, which is an irrigation water source and is affected significantly by the seasons. Summer is a dry season and not suitable for excitation experiments. However, 2016 was a special year. The water level in the dry season of summer was still above 15m, and was suitable for excitations throughout the year. From June 6, 2017, affected by the water level, the experiment was stopped and excitations were resumed at the end of September. From January 1st, 2016 to June 6th, 2017, there were 5 fixed stations in the experimental area of the West Yunnan (the seismometers mainly include BBVS-60 and KS2000, the data collector is EDAS-24IP), and 40 active source receivers (the seismometer was Guralp-40T, and the data collector was Reftek130B). The data was recorded and collected (the data used in this paper is concentrated in a short period of time, with stable GPS timing). The collection shows that all the excitation data can be obtained from January 1, 2016 to June 6, 2017 except for the airgun maintenance and troubleshooting. The daily average number of excitations are kept at around 13. However, data was missing at individual stations because of instrumental troubles. According to the regional characteristics, 13 out of 45 stations with high signal-to-noise ratios and complete data are selected (including 3 fixed stations and 10 active source stations). Airgun source receiving stations, excitation source and the epicentral distribution of the two earthquakes are shown in Fig. 1.
The research shows that the waveform acquired by the airgun source excitation in the water contains the information of the underground medium, and also carries the interference information formed by the water body, the water level and the excitation condition. To eliminate interference information, Wang Baoshan et al. (2008) explored the wake interference technique to process the airgun source waveform. The wake-wave interference technique approximates the near-field record excited by the airgun source as the source time function, and deducts the source characteristics in the far-field recorded waveform by deconvolution, thereby obtaining the Green's function from the source to the station (Wang Baoshan et al., 2012; Zhai Qiushi et al., 2016). For the approximate Green's function, the waveform cross-correlation method is used to obtain the high-precision travel time change of different phases.
The airgun console generates an excitation log during the airgun excitation process, recording the excitation time, but due to the influence of the gun control equipment clock calibration, mechanical delay, time recording accuracy, etc., the excitation time and the real excitation time in the gun control log have large difference (Zhou Qingyun, 2018). Therefore, using the CKT0 near the airgun source (about 50m from the airgun excitation source) as the reference station, intercepting a certain excitation recording waveform signal as the original signal template, and using the cross-correlation processing technique to perform cross-correlation processing on each excitation recording waveform of the reference station help determine the relative airgun excitation moments (Zhang Yuansheng et al., 2017). On the basis of obtaining the excitation time, 13 stations are selected to intercept the recorded waveform data with a duration of 100s based on the excitation time.The intercepted waveform data of the reference station (CKT0, the interception duration is 5s) and the 13 stations are de-averaged, de-tilted, de-dc processed, and 3-5Hz Butterworth bandpass is filtered, noise interference is reduced and the distorted recorded waveform is manually deleted. Then, the data of each station of the same gun is deconvolved with the reference station data, and the influence of the source is removed when the alignment is obtained, and the Green's function from the station to the reference station is obtained. On this basis, all the Green's functions of each station are superimposed, and the superposition results are used as the reference waveform templates of the respective stations, and the Green's function of each station is superimposed on a daily basis. Finally, the superposition result of each day is cross-correlated with the reference template. The time delay corresponding to the maximum correlation coefficient is the travel time of the two similar waveform windows. In order to improve the accuracy, the cosine interpolation of the calculation result is extracted to obtain the travel time change of higher precision (Liu Zifeng et al., 2015).1.3 Consistency Analysis of the Excitation Waveform of Airgun Source
The consistency of the airgun source is its most essential feature and outstanding advantages (Chen Yong et al., 2007c). It is the basic requirement for the active source "4-D" seismology research, the construction of the "underground cloud map", and the superimposition of multiple records to improve the signal-to-noise ratio of the data (Wang Bin et al., 2015). Measuring small changes in wave speed requires a good repeatability of the selected waveform. Let us take the EY16 station far away from the excitation source as an example (the EY16 is about 110km away from the excitation source, and the waveform record is relatively complete during the study period). The consistency of the excitation waveform of the airgun source is tested. The detection situation is shown in Fig. 2. Fig. 2(a) is a reference template, Fig. 2(b) is the superimposed waveform of 13 gun data on March 20, 2016 (single-day superposition result), and Fig. 2(c) is a correlation coefficient of waveform cross-correlation analysis between single-day superimposed data and a reference template. It can be seen from Fig. 2 that the main phase correlation coefficients are all around 1.0, which indicates that the waveform generated when the airgun source is repeatedly excited at the same place is highly repetitive.
Zhou Qingyun et al. (2018) used the data excited by the Binchuan seismic signal launching station in February, 2017 to systematically study the travel time change characteristics of airgun signals under different excitation pressures, sinking depths and horizontal displacement conditions. The results indicate that because the airgun signal changes in different time windows, it is difficult to eliminate the influence of pressures by the technical means. Therefore, it is recommended to fix the excitation conditions in the future airgun excitation, and use the existing data to select the similar amplitude data in the related travel time change research. The Binchuan airgun source excitation system consists of four Bolt Longlife airguns with a single gun capacity of 2000in3. The airgun sinking depth is about 10m below the floating platform, and the four airguns generate about 8.91×106J of energy, which is equivalent to one ML0.7 Natural earthquakes. The excitation pressure and the airgun sinking depth are basically unchanged during the daily excitation. Therefore, in the case of the fixed excitation pressure and sinking depth, we select a travel time from a similar seismic phase with the largest correlation coefficient and the maximum amplitude (as shown by the dotted line in Fig. 2). The length of the time window is determined in principle to fully represent a seismic phase with a value between 0.5s-1.0s. There are small differences in the values of different stations.2 CHARACTERISTICS AND STRESS ANALYSIS OF WAVE VELOCITY BEFORE AND AFTER TWO EARTHQUAKES
The waveform data of 13 stations were processed by the methods in 1.2 and 1.4 (see Table 1 for the distance between each station and the excitation source and the two earthquakes), and the travel time curves of the respective stable phases were extracted. Combined with the GNSS baseline data (GNSS site distribution and related baseline changes shown in Figs. 1 and 3), the characteristics of wave velocity changes before and after the two earthquakes were analyzed. According to the change characteristics of each station's travel time, combined with the location of the station, we divide 13 stations into 4 areas. The area division is shown in Fig. 1.
It can be seen from Fig. 1 that every three stations in the regions (1), (2) and (3) are roughly in a straight line, and the distance from the excitation source is from near to far, and the respective stable phase transition curves are very similar. The station travel time changes are shown in Figs. 4, 5 and 6. Judging from the travel time changes, those of the YUL, CHT and 53039 stations have relatively large amplitudes with the maximum up to ±0.03s. The other stations vary considerably. According to the variation characteristics of each station's travel time change and the occurrence time of the two earthquakes, we divide the period from January, 2016 to June, 2017 into eight.
Period 1: from January 4 to February 22, 2016. During this period, the water level of the reservoir was relatively stable and the water level maintained at around 22.5m. The travel time change of the 53256 station in the region (1) was stable, and that of the EYA station had small fluctuations. The EY16 station was set up in February, and data was recorded from February 15 on. The travel time change of the 53036 station in the region (2) was stable and that of the 53279 station had slow rise with a small amplitude. The YUL station had an obvious lifting process with a variation range about 0.025s. The travel time changes of the three stations in the region (3) were relatively stable.
Period 2: from March 18 to April 26, 2016. During this period, the reservoir water level decreased by 1.59m. The travel time of the stations of 53256, EYA, EY16, 53279, 53036, YUL, and 53258 in the three regions showed a trend of simultaneous uplift, and the YUL station had the largest change in amplitude about 0.02s. The travel time changes of the stations CHT and 53039 in the region (3) were relatively stable.
From January 4 to April 26, 2016, the baselines of Shidian-Yunlong (Fig. 3(a)), Zhongdian-Yunlong (Fig. 3(b)), and Xiaguan-Yunlong (Fig. 3(c)) related to the Yunlong GNSS stations showed different stress characteristics. The baseline of Zhongdian-Yunlong showed a trend of extrusion enhancement. The Shidian-Yunlong and Xiaguan-Yunlong baselines fluctuated from January to March. In mid-March, Shidian-Yunlong began to squeeze, and the Xiaguan-Yunlong baseline showed a pulling trend. The baseline results show that in Period 1, the stress accumulation in other regions is not obvious except for the YUL, which represents the region (2). Therefore, the travel time in the other stations except the YUL and 53279 changes smoothly. In Period 2, the stress extrusion characteristics of the regions (1) and (2) are more obvious. The travel time change of most stations is significantly affected by the regional stress, showing the synchronous travel time rising characteristics. The CHT and 53039 stations in the region (3) are affected by the Xiaguan-Yunlong baseline tensile characteristics (the regional stress accumulation is not obvious), and the travel time changes are relatively stable. During Period 2, the water level dropped by 1.59m. The CHT and 53039 travel time changes unaffected by the regional stress did not show the same or opposite trend with the water level change. It can be seen that the water level drop of 1.59m did not have a substantial impact on the travel time change of each station.
Period 3: from May 6 to June 4, 2016 (before and after the Yunlong MS5.0 earthquake on May 18). The water level dropped by 3.64m during this period. Before and after the Yunlong earthquake, the baselines of Shidian-Yunlong, Zhongdian-Yunlong and Xiaguan-Yunlong showed synchronous fluctuations, indicating that the regional stress was adjusted by the influence of the Yunlong earthquake process and the travel time changes of the nine stations in the three regions showed fluctuations. The YUL station had the largest fluctuation with the variation interval of ±0.025s.
Period 4: from June 18 to July 31, 2016. After June 4, the water level of the reservoir continued to drop, and reached the lowest level of 14.25m in the entire year on June 10. At this time, the staff repaired the airguns and temporarily stopped the excitation, so there was no observation data during the period, and it was restored on June 18. After that, the water level was on the rise since then. During this period, the relevant baseline fluctuations have decreased. Correspondingly, the 53256, EYA, EY16, 53036, YUL, CHT, and 53039 travel time changes still had different degrees of fluctuation, of which the YUL station fluctuation was the most obvious, the maximum fluctuation ranged between ±0.03s. The 53258 and 53279 stations had a small amount of travel time drop.
Period 5: from August 10 to November 12, 2016. During this period, the water level increased by 4.13m, and the 53256, EY16, 53258, and 53039 stations in the regions (1) and (3) changed smoothly. The EYA station experienced slight fluctuations from September 14 to November 12. The 53279, 53036 and YUL stations in the region (2) and the CHT station in the region (3) can be analyzed in two periods: From August 10 to September 18, the four stations had relatively stable changes. From September 27 to November 12, the 53279, 53036, and YUL stations had a synchronous upward trend, and the YUL station was the most obvious, with a variation of about 0.02s. The CHT station had a tendency to decline in travel time, and the amount of change was about 0.02s. In combination with Fig. 3, from September 27 to November 12, the Zhongdian-Yunlong and Xiaguan-Yunlong baselines had synchronous fluctuations, while the Shidian-Yunlong baseline was relatively stable, and the travel time anomalies of the 53279, 53036, YUL and CHT stations might be related to the stress adjustment process in the small area represented by the YUL and CHT stations in a short-term.
Period 6: from December 19, 2016 to February 19, 2017. The water level dropped by 2.05m during this period. Except the 53256 and 53039 stations, the other seven stations had different degrees of travel time descent, and the 53036, EYA, and YUL stations (distances from the earthquake: 10km, 29km, and 43km, respectively) had most obvious changes between 0.015s and 0.02s. As can be seen from Fig. 3, before the Yangbi Earthquake, the Xiaguan-Yunlong baseline showed large fluctuations away from the long trend, which indicated that the stress field in the pre-earthquake region was adjusted, resulting in an accelerated wave velocity of the airgun source signal received by the adjacent stations.
Period 7: from March 13 to April 6, 2017 (before and after the Yangbi MS5.1 earthquake). The water level dropped by 0.4m during this period. Due to the impact of the Yangbi earthquake, the relevant baselines had varying degrees of fluctuations. The data of the 53279, 53036, 53039 stations were missing due to the instrument failures. The remaining six stations had varying degrees of fluctuations, but the overall fluctuation was small.
Period 8: from April 20 to June 6, 2017. During this period, the water level changed significantly, declining by 4.96m. The 53256, EYA, and EY16 stations in the (1) area (the distances of the three stations from the earthquake were: 49km, 29km, and 44km respectively) they had a tendency to rise synchronously, and the EYA station was the most obvious, with a variation of about 0.02s. The other two stations varied by approximately 0.01s. The travel time of the YUL station in the region (2), the 53258 and CHT stations in the region (3) had a synchronous downward trend (the distances of the three stations from the Yangbi earthquake were 43km, 85km, and 49km respectively), and the YUL and CHT had the most obvious changes (the distance of the two stations from the Yangbi earthquake was about the same at 45km), and the variation was about 0.02s. Referring to Fig. 1, it can be found that the 53256, EYA, and EY16 stations in the (1) region are located on the north side of the Yangbi MS5.1 earthquake, while the YUL, 53258, and CHT stations in the regions (2) and (3) are located on the south side of the Yangbi earthquake. Combined with the GNSS baseline data in Fig. 3, it can be seen that in the short-term, the Zhongdian-Yunlong baseline has compressive enhancement, and the Shidian-Yunlong and Xiaguan-Yunlong baselines have a tendency to increase in tension. The three baseline changes indicate that affected by the Yangbi MS5.1 earthquake, the stress characteristics of different natures appeared on both sides of the earthquake in the short-term. This is consistent with the fact that the travel times of the 53256, EYA, and EY16 stations become larger and those of the YUL, 53258, and CHT stations become smaller.
From January 2016 to June 2017, the reservoir water level underwent two major changes. From January 1, 2016 (water level was 22.7m) to June 10 (water level was 15.25m), the reservoir water level decreased by 8.45m in total. The water level returned to the original height (22.7m) on November 3. From November 5, 2016 to June 6, 2017 (water level was 13.74m), the reservoir water level again presented a downward trend, dropping by 8.96m in total. Of the 9 stations in three regions, 7 stations (the 53258 and 53279 stations excluded) have their distance from the excitation source greater than 45km from the excitation source. As for the relationship between the travel time change and the reservoir water level change in the 2 large-scale water level change process, the travel time change of the 9 stations does not follow the water level change and has a stable trend. Some stations have the phenomenon that when the water level becomes stable, travel time change becomes large, and when the water level change becomes dramatic, the travel time change becomes stable. It can be seen that the change of reservoir water level does not have a substantial impact on the travel time change of the 9 stations for three regions. In contrast, the regional stress adjustment process in different time domains before and after the earthquake is more likely to have a greater impact on the travel time of the airgun source signals received by each station.2.2 Characteristics and Stress Analysis of Wave Velocity in the Region (4)
The distance between the four stations in the region (4) and the excitation source is less than 25km (except for the 53266 of 24km, and the others less than 10km). The epicenter distance from the Yunlong MS5.0 earthquake is greater than 90km, and the epicenter distance from the Yangbi MS5.1 earthquake is greater than 60km. The four stations recorded clear excitation wave, high signal-to-noise ratio, stable travel time change and small fluctuations. The travel time change of the three stations in the four stations was between ±0.01s and the 53272 station has a slightly higher one between ±0.015s. The travel time change of each station is shown in Fig. 7.
Combined with the change of water level, it can be analyzed in 6 periods: From January 4 to February 22, 2016, the water level of the reservoir was stable and the water level dropped by 0.67m. In this period, the four stations had a stable travel time with the variation less than 0.001s; From March 18 to June 4, the water level of the reservoir changed significantly, and the change rate went from slow to fast. On June 10, it reached the lowest value in the whole year. The water level change reached 8.05m. The travel time of the 4 stations showed a synchronous upward trend from the water level, and the rate of change went from slow to fast. The increased amount of the travel time of the four stations in the short-term is more than 0.007s, with the 53272 station as the most significant about 0.014s. After the Yunlong MS5.0 earthquake, there were small fluctuations in the change of travel time. From June 17 to August 22, the water level of the process recovered rapidly, the water level rise value was 5.87m, and the travel time of the four stations showed a synchronous downward trend. In the short term, the amount of change exceeds 0.004s, and the 53272 had the largest change with a falling value of 0.008s. The travel time changes of the four stations maintain a good correlation with the water level change. When the water level is stable, the travel time is also stable. When the water level rises rapidly, the travel time decreases rapidly. The difference is that in the middle and late July, there was a small reverse rise in the short-term. From August 22 to November 24, the travel time change of the four stations in this period showed a good similarity with the change of water level. The water level of the whole process increased by 2.14m, and the rise of the travel time change of the four stations was around 0.006s. The 53272 station had the maximum of 0.010s. Between September 18 and September 27, the 53266, 53272 and 53274 stations had a more obvious process of travel time declining. From December 19 to January 25, the water level was stable, and the travel time change of the four stations was smooth. After that, the water level descended in a stepwise manner until April 8, 2017. During this process, the travel time change of the four stations did not increase or decrease synchronously. The travel time of the 53266 and 53272 went up, with the 53272 as the most obvious of 0.009s. The travel time of the 53262 and 53274 stations went down. After the Yangbi MS5.1 earthquake, there were small fluctuations. From April 17, 2017 to June 6, 2017, the water level dropped by 5.12m. With the rapid decline of the water level, the travel time change of the four stations increased in a synchronous manner. Among them, the 53272 had the most obvious upward trend, reaching 0.009s.
Based on the GNSS baseline data, the travel time changes of the four sites in the region (4) are analyzed. As can be seen from Fig. 1, the Xiaguan-Yongsheng baseline spans over the region (4). Therefore, the baseline change can quantitatively characterize the local stress variation of the region. It can be seen from Fig. 3(d) that from January 4 to August 22, 2016, the Xiaguan-Yongsheng baseline showed abnormal fluctuations, but the overall trend was tensile. During this period, the regional stress accumulation was not obvious, and the travel time changes of the four stations maintained a good negative correlation with the water level changes. From August 22 to November 24, the Xiaguan-Yongsheng baseline had a process of pressure enhancement. There was a slight fluctuation in mid-September, due to the regional crushing stress, and the original negative correlation between the travel time changes of the four stations and the water level was broken, and the change in travel time showed a synchronous upward trend, and fluctuated in mid-September. At the end of November, the Xiaguan-Yongsheng baseline showed a rapid extension phenomenon, and it recovered rapidly at the end of December. During this period, the travel time changes of the four stations returned to be stable. At this time, the water level was stable and the two corresponded with each other. From the beginning of January 2017, the Xiaguan-Yongsheng baseline showed a rapid exponential trend and recovered rapidly at the end of February. The negative correlation between the travel time change and the water level change of the 53266 and 53272 stations recovered, but the travel time change of the 53262 and 53274 stations had the similar step-down feature with the water level change which may be related to the rapid adjustment of the regional stress in the short term. The extrusion or tensile characteristics of different regions result in different characteristics of the four stations. Beginning in mid-April, the baseline anomaly recovered and showed tensile characteristics, and the travel time change in the four stations and the change in water level returned to the original negative correlation. It can be seen from the above process that the regional stress adjustment is more likely to affect the travel time change of the four stations than the water level change.3 CONCLUSIONS AND DISCUSSION
This paper used the airgun source signals recorded from 13 stations in 4 regions from January 2016 to June 2017 and the cross-correlation detection technology to obtain the stable phase-seismic variation characteristics of each station. The Yunlong MS5.0 and Yangbi MS5.1 earthquakes are taken as a sample for the in-depth analysis of the characteristics of wave velocity changes before and after the earthquake, combined with GNSS baseline data. The results show that:
(1) The travel time changes of 13 stations in 4 regions are greatly affected by the impact of the two earthquakes. Before the Yunlong MS5.0 earthquake, the travel time changes of the 53256, EYA, EY16, 53279, 53036, YUL, and 53258 stations began to rise for about 40 days in the first two months before the earthquake, with a maximum change of 0.02s (the YUL station). Before the Yangbi MS5.1 earthquake, the travel time changes of the 53279, 53036 and YUL stations began to rise for about 45 days in the first 5 months before the earthquake. The maximum change was 0.02s (the YUL station). The difference was before the Yangbi MS5.1 earthquake, the travel time changes of the relevant stations had a decline in 3 months. In contrast, the Yunlong MS5.0 earthquake had a greater impact on the travel time changes of related stations in the adjacent area. Due to the influence of the regional stress adjustment before and after the Yangbi earthquakes, all stations had different amplitude fluctuations before and after the two earthquakes. The fluctuation amplitude and duration of each station in different regions were different, but the fluctuations amplitude before and after the Yangbi earthquake were small. Before and after the two earthquakes, the abnormal characteristics and fluctuations of the YUL, EYA and 53036 stations were the most obvious. The above-mentioned travel time anomalies have good correspondence with the short-term regional stress characteristics characterized by the GNSS baseline data. It is worth noting that the epicenter distances of the EY16 station from the Yunlong MS5.0 and Yangbi MS5.1 earthquakes were 9m (ranked 1st) and 44m (ranked 4th), but the travel time rise during the corresponding period and the fluctuation characteristics before and after the earthquakes were relative. It was not obvious at other stations. This may be due to the stable structure of the EY16 station, which is relatively less affected by earthquake incubation, occurrence and post-earthquake stress adjustment. The specific reasons need further study.
(2) Compared with the regions (1), (2) and (3), the region (4) is farther from the epicenter of the two earthquakes (more than 60km), and the time difference of the four stations is less than 0.01s. The magnitude of the change was small, but there were slight fluctuations in the short term before and after the two earthquakes. There is a certain degree of negative correlation between the change of travel time in the region (4) and the change of water level. In the short term, with the adjustment of regional stress, the negative correlation is broken. The travel time change and water level change in other regions are not obvious. The water level change does not have a substantial impact on the change of travel time. This is because the data processing in the previous stage has used the deconvolution of the reference stations to eliminate the influence of the excitation source. In contrast, the adjustment of the regional stress field is more likely to have a substantial impact on the time-dependent changes of the relevant stations.ACKNOWLEDGEMENTS
The cross-correlation delay calculation program of this paper was provided by Professor Wang Baoshan at the University of Science and Technology of China, and the anonymous reviewer proposed valuable revisions. I would like to express my heartfelt gratitude to the above people and reviewers for their valuable suggestions.
Chen Jia, Li Xiaobin, Yang Jun, Ye Beng. Large volume air-gun source spectrum characteristics of Binchuan, Yunnan[J]. Earthquake Research in China, 2016, 32(2): 216-221 (in Chinese with English abstract).
Chen Yong, Zhu Rixiang. Proposed project of "Underground Bright Lump"[J]. Advances in Earth Science, 2005, 20(5): 485-489 (in Chinese with English abstract).
Chen Yong, Wang Baoshan, Ge Hongkui, Xu Ping, Zhang Wei. Proposed of transmitted seismic stations[J]. Advances in Earth Science, 2007a, 22(5): 441-446 (in Chinese with English abstract).
Chen Yong, Zhang Xiankang, Qiu Xuelin, Ge Hongkui, Liu Baojin, Wang Baoshan. A new way to generate seismic waves for continental crustal exploration[J]. Chinese Science Bulletin, 2007b, 52(16): 2264-2268. DOI:10.1007/s11434-007-0247-4
Chen Yong, Li Yijing. Seismic wave radar research:using active source to detect continental crust structure[J]. Journal of University of Science and Technology of China, 2007c, 37(8): 813-819 (in Chinese with English abstract).
Chen Yong, Wang Baoshan. Study on the underground structure and its changes in the shallow part of continent with an Airgun source. In: Proceedings of the Annual Meeting of the Chinese Geophysical Society. Beijing: Chinese Geophysical Society, 2016 (in Chinese).
Chen Yong, Wang Baoshan, Yao Huajian. Seismic airgun exploration of continental crust structures[J]. Science China Earth Sciences, 2017, 60(10): 1739-1751. DOI:10.1007/s11430-016-9096-6
Li Xiaobin, Ye Beng, Yang Jun, Chen Meng, Jin Mingpei, Wang Baoshan. Research on comparison of excitation efficiency of air guns with various combinations in water reservoir[J]. Journal of Seismological Research, 2016, 39(3): 458-465 (in Chinese with English abstract).
Liu Zifeng, Su Youjin, Wang Baoshan, Wang Bin, Yang Jun, Li Xiaobin. Study on analysis method of travel time variations of seismic wave of active source in Binchuan[J]. Journal of Seismological Research, 2015, 38(4): 591-597 (in Chinese with English abstract).
Luan Yi, Yang Hongfeng, Wang Baoshan. Large volume air-gun waveform data processing (I):Binchuan, Yunnan[J]. Earthquake Research in China, 2016, 32(2): 305-318 (in Chinese with English abstract).
Okaya D., Henrys S., Stern T. Double-sided onshore-offshore seismic imaging of a plate boundary:"super-gathers" across South Island, New Zealand[J]. Tectonophysics, 2002, 355(1/4): 247-263.
Qiu Xuelin, Chen Yong, Zhu Rixiang, Xu Huilong, Shi Xiaobin, Ye Chunming, Zhao Minghui, Xia Shaohong. The application of large volume airgun sources to the onshore-offshore seismic surveys:implication of the experimental results in northern South China Sea[J]. Chinese Science Bulletin, 2007, 52(4): 553-560. DOI:10.1007/s11434-007-0051-1
Wang Bin, Wu Guohua, Su Youjin, Wang Baoshan, Ge Hongkui, Jin Mingpei, Yang Jun, Zhang Qiming, Li Tao. Site selection and construction process of Binchuan earthquake signal transmitting seismic station and its preliminary observation result[J]. Journal of Seismological Research, 2015, 38(1): 1-6 (in Chinese with English abstract).
Wang Bin, Li Xiaobin, Liu Zifeng, Yang Jun, Yang Runhai, Wang Baoshan, Yang Wei. The source and observation system of Binchuan earthquake signal transmitting seismic station and its preliminary observation results[J]. Earthquake Research in China, 2016, 30(3): 316-325.
Wang Baoshan, Zhu Ping, Chen Yong, Niu Fenglin, Wang Bin. Continuous subsurface velocity measurement with coda wave interferometry[J]. Journal of Geophysical Research:Solid Earth, 2008, 113(B12): B12313. DOI:10.1029/2007JB005023
Wang Baoshan, Yang Wei, Wang Weitao, Yuan Songyong, Song Lili, Ge Hongkui, Wang Bin, Wu Guohua, Su Youjin, Liu Xuejun, Jin Mingpei, Yang Jun. Un-blur the signature of large volume airgun excited in a reservoir with deconvolution. In: Proceedings of the Twenty-eighth Annual Meeting of the Chinese Geophysical Society. Beijing: Chinese Geophysical Society, 2012 (in Chinese).
Wang Baoshan, Ge Hongkui, Wang Bin, Wang Haitao, Zhang Yuansheng, Cai Huiteng, Chen Yong. Practices and advances in exploring the subsurface structure and its temporal evolution with repeatable artificial sources[J]. Earthquake Research in China, 2016, 30(3): 284-297.
Wang Weitao, Wang Baoshan, Ge Hongkui, Chen Yong, Yuan Songyong, Yang Wei, Li Yijin. Using active source to monitor velocity variation in shallow sediment caused by the Wenchuan earthquake[J]. Earthquake Research in China, 2009, 25(3): 223-233 (in Chinese with English abstract).
Wang Weitao, Wang Baoshan, Jiang Shengmiao, Hu Jiupeng, Zhang Yuansheng. A perspective review of seismological investigation on the crust at regional scale using the active Airgun source[J]. Journal of Seismological Research, 2017, 40(4): 514-524 (in Chinese with English abstract).
Wang Weitao, Wang Baoshan, Jiang Shengmiao, Hu Jiupeng, Zhang Yuansheng. Regional scale seismological investigation on the continent crust using airgun sources——A perspective review[J]. Earthquake Research in China, 2018, 32(3): 315-329.
Wei Yunyun, Wang Haitao, Su Jinbo, Chen Xiangjun, Wang Qiong. The preliminary study on travel time abnormal variation of reflection wave phase of air-gun in Xinjiang before two earthquakes with MS5.0[J]. Earthquake Research in China, 2016, 32(2): 270-281 (in Chinese with English abstract).
Yang Wei, Ge Hongkui, Wang Baoshan, Yuan Songyong, Song Lili, Jia Yuhua, Li Yijin. Velocity changes observed by the precisely controlled active source for the Mianzhu MS5.6 earthquake[J]. Chinese Journal of Geophysics, 2010, 53(5): 1149-1157 (in Chinese with English abstract).
Zhai Qiushi, Yao Huajian, Wang Baoshan. Study on the deconvolution method and processing flow of airgun source data[J]. Earthquake Research in China, 2016, 30(3): 394-404.
Zhang Yuansheng, Wang Baoshan, Chen Yong, Wang Lanmin, Zou Rui, Qin Manzhong, Guo Xiao, Shen Xuzhang, Wei Congxin, Liu Xuzhou, Wang Yahong, Sun Dianfeng, Guo Yingxia, Yin Liang. Travel-time variations before and after two major earthquakes derived from active-source seismic data[J]. Chinese Journal of Geophysics, 2017, 60(10): 3815-3822 (in Chinese with English abstract).
Zhang Yunpeng, Li Xiaobin, Wang Weitao, Wang Weitao, Wang Baoshan, Ye Beng, Yang Jun, Wang Bin. Data service system and data quality assessment for mobile observation of transmitting seismic stations in Binchuan, Yunnan[J]. Journal of Seismological Research, 2017, 40(4): 525-533 (in Chinese with English abstract).
Zhou Qingyun. Preliminary study on directivity of Binchuan air-gun source energy radiation[J]. Journal of Geodesy and Geodynamics, 2018, 38(7): 754-759 (in Chinese with English abstract).
Zhou Qingyun, Chen Junlei. Influence of different triggering conditions of airgun source on travel time changes[J]. Journal of Seismological Research, 2018, 41(2): 264-272 (in Chinese with English abstract).