| Citation: |
Baojun Yin, Li Ma, Huizhong Chen, Jianping Huang, Chaojun Zhang, Wuxing Wang (2009). Characteristics of coseismic water level changes at Tangshan well for the Wenchuan MS8.0 earthquake and its larger aftershocks. Earthq Sci 22(2): 149-157. DOI: 10.1007/s11589-009-0149-4
|
The Wenchuan MS8.0 earthquake occurred at 14:28:04 on 12th May 2008 at the southeast part of the Longmenshan fold zone, Sichuan province, China. After the main shock, there occurred 15 larger aftershocks (MS≥5.4) in a 200 km long belt within the fold zone. The 16 earthquakes from the same seismic-bearing fold zone, caused a series of coseismic changes at Tangshan well, and the data is very precious for further research on water level coseismic phenomena.
Some researchers have discussed changes of water level and water temperature at Tangshan well (Cao, 1988; Zhang et al, 1998, 2002; Che and Yu, 2004; Shi et al, 2007; Ma et al, 2008). Zhang et al(2005b, 2007) compared coseismic recording ability of analog record with that of digital record, and suggested that there exists distortion for coseismic recording of the digital record since the sampling rate of the digital record is one minute whereas the period of the coseismic recording is by second; however, the analog water level record has the advantage for recording the amplitude and the duration of coseismic water level oscillation. As for research on coseismic water level variation to large earthquakes, some researchers analyzed coseismic water level variation characteristics and response mechanisms of Chi-Chi (Jiji) earthquake and Sumatra earthquake using the data from several wells (Huang et al, 2000; Liu et al, 2005; Zhang et al, 2005a; Kitagawa et al, 2006). However, there is no statistical coseismic research on a well responding to an earthquake sequence to the best of our knowledge. Tangshan well provides us the possibility due to its high sensitivity.
The paper uses the analog water level record at the Tangshan well to analyze time and amplitude parameters of coseismic variation caused by the Wenchuan main shock and its 15 larger aftershocks. The statistical correlation between seismic parameters and coseismic amounts is calculated, and the possible mechanism of coseismic changes is discussed.
The Tangshan well is located in Dazhao park in Tangshan city, which is also called Tangshan mine or Shanxi No.2 well. Dazhao park is inside Tangshan city, the altitude of which is 25.43 m. The well was drilled on October 30 in 1969 with depth 286.6 m. The hole was sealed at 207 m depth, with casing depth to 154 m. The aquifer layer consists of Ordovician 1imestone of about 50 m thick. The well lies between Yanshan fold zone and North China plain subsidence zone, with Tangshan fault in NE direction and a buried fault in WNW direction going through the well. The average burying depth is about 60 m now.
The SW40 nilometer was used to observe the ground water level from January 1981. It was upgraded to SW40-1 nilometer in 1993. Then in September 2001, a LN-3 digital nilometer was set up with the analog nilometer runing simultaneously. SW40-1 nilometer is a mechanical apparatus with buoy to detect water level changes. Its mechanical clock drives an ink pen to record the water level changes on the paper around the cylinder. There are staff members replacing the starting position of the pen or changing a new piece of paper at 8 am everyday. The precision of the analog record is 1 mm, and the daily time error is about two minutes. Because the sampling rate of LN-3 digital nilometer is one minute and the period of coseismic changes is measured by second, it is impossible to record coseimic changes completely. Therefore, we use analog data to examine the coseismic changes of the 16 earthquakes (Figure 1).
There are prominent annual tendency, earth tides and barometric effects in the water level at the Tangshan well. However, comparing with the period of seismographs, the period of these effects are much larger. Therefore, these effects have little influence on the water level changes during the short coseismic oscillation. Hence the error of using the original data to examine coseismic changes is very small. The effect on the statistical results will not be significant.
Douhe seismic station is located on the base rock of the south hill of Douhe reservoir which lies at the northeast of Tangshan city, about 16.3 km far away from the Tangshan well. The altitude of the station is 55.42 m. The 3-components digital broad-band seismometer has a responding characteristic of 0.05-20 Hz. The dynamic range is about 120 dB. The digitalizer EDAS-C24 is made by Geodivice Inc., China. The sampling rate is 50 Hz, and the timing precision is ≤1 ms.
The Wenchuan earthquake occurred on the southwest section of the Longmenshan fold zone. From northwest to southeast, The Longmenshan fold zone can be divided as follows: the Houshan fault extending along Wenchuan-Maowen direction, the Zhongyang fault extending along Yingxiu-Beichuan direction, and the Qianshan fault extending along Dujiangyan-Anxian direction. The three faults immerged into one shearing zone at the depth about 20−30 km (Wang et al, 2008). After the main shock, from 12th May to 27th May, there occurred 11 aftershocks (MS≥5.4) in the main shock region. Then from 24th July to 5th August, four aftershocks (MS≥5.4) occurred in the northeast Qingchuan region. Most of the earthquakes are thrust-fault earthquakes. The focal mechanism of most of the earthquakes near Lixian is strike-slip, whereas there are both strike-slip and thrust-fault earthquakes in the east of Qingchuan (Hu et al, 2008).
The basic parameters of the 16 earthquakes (MS≥5.4) are listed in Table 1. Except that MW magnitude and depth are from Harvard university CMT catalogue, all the others are from China Earthquake Networks Center (CENC) and Shaanxi earthquake networks (ML magnitude) on the website of China Earthquake Administration. The well-epicenter distance is calculated using parameters from CENC catalogue. The arrival times of body waves and maximum amplitudes of Love and Rayleigh waves of the 16 earthquakes recorded at Douhe seismic station are also listed.
|
DownLoad:
CSV
The typical feature of coseismic water level oscillation at Tangshan well to the Wenchuan earthquakes is as follows. When seismic wave arrives, there is a pulse change in water level; then the water level continues oscillating, and has a water level step in the process (Figure 1). About 3 minutes after the Wenchuan MS8.0 earthquake, the water level at the Tangshan well emerged a short oscillation for about two minutes with maximum amplitude of 19 mm. After comparing with the theoretical phase arrival times calculated from the program on NEIC website, we find that the beginning time of the oscillation is nearly at the same time with the arrival times of the Pn and P waves. The amplitude of the oscillation keeps increasing, and reaches its peak value (2.434 m, from digital record) in 10 minutes. The duration of the oscillation is about 240 minutes, and there appears an 11 mm step change in total. As for the other 15 larger aftershocks, if the magnitude is around MS6.0, there is always a pulse with maximum amplitude at first, then following an oscillation with decreasing amplitude, accompanied by a step (See EQ #3 in Figure 1). If the aftershock is around MS5.5, there always occurs a swift step change at first; then the water level keeps its changing tendency until the time when the water level arrives at the place where it should be determined according to its original tendency, accompanied by minor oscillations. We have not found significant coseismic changes which can be distinguished from the background noise in the water level for earthquakes with MS < 5.4 in the analog record at Tangshan Well.
Table 1 lists maximum amplitudes of coseismic water level changes caused by the Wenchuan earthquakes, the time between seismic phase arrivals and the maximum amplitude of the coseismic oscillations, the amount of step change, and duration of oscillations, etc. (The coseismic changes of EQ #2 appears to be overlapped by that of the main shock), the time of maximum coseismic amplitude comparing to seismogram phases can be divided into four types, P-S, S-Qm, Qm-Rmz, > Rmz, they are numbered to 1−4 correspondingly, for drawings in Figures 3−6.
Some researchers have studied correlation between earthquake magnitude and well-epicenter distance for coseismic water level changes. Matsumoto (1992) analyzed the data from Haibara well in Japan and obtained the relationship between the magnitude and well-epicenter distance M=2.45lgD+0.69. Kitagawa and Matsumoto (1996) analyzed the water level at a well in Tokai area, Japan, and suggested a relationship as M=2.62lgD. King et al (1999) provided a relationship of M=2.5lgD for Tono well, Japan. Yang et al (2005) analyzed coseismic phenomena at Dazhai well, Simao county, Yunnan province in China, and suggested an equation as M=1.176lgD+3.104. These studies used different earthquake catalogues and different types of magnitude scaling. The results differ from each other as well. Here, we use three types of magnitude scaling of the Wenchuan earthquakes to examine the relationship between the magnitude and well-epicenter distance for the coseismic changes at the Tangshan well.
We use MS, ML and MW scaling of the earthquakes listed in Table 1 which have caused coseismic water level changes at the Tangshan well, to examine the relationship between the magnitude and well-epicenter distance (Figure 2). Comparing the three relationships we obtained, the result using MS magnitude from CENC is similar to the result using MW magnitude from Harvard, with nearly the same slope, though the fitted line of using MS is slightly higher than that of using MW magnitude as shown in the figure. The result using ML magnitude from Shaanxi networks has a remarkable difference from the other two. The same conclusion is drawn when calculating the lower limit of the relationship between the earthquake magnitude and well-epicenter distance for coseismic changes. Comparing the equations for the Tangshan well with that of other wells mentioned above, if we use the same well-epicenter distance to calculate the lowest magnitude at which an earthquake can cause coseismic changes at a well, the Tangshan well needs the lowest magnitude to be triggered to oscillate, indicating that the Tangshan well is more sensitive.
Figure 3 shows the correlations between three types of magnitude scaling and coseismic parameters for the Tangshan well. For the four coseismic parameters, the strongest correlations are between the maximum amplitude and the three types of magnitude scaling, and between the amount of step change and the magnitudes, the all is larger than 0.85. The correlations between the duration of coseismic changes and the magnitudes are all larger than 0.8. However, the time when coseismic oscillation reaches its maximum amplitude has less correlation with the magnitudes. Comparing the correlations respectively for the three types of magnitudes, MS magnitude always has the strongest correlation with the coseismic parameters except for the time when coseismic oscillation reaches its maximum amplitude. The results for MW magnitude are close to that of MS magnitude, whereas ML magnitude always has the weakest correlation.
Figure 4 shows the correlations between well-epicenter distance and parameters of coseismic changes caused by the Wenchuan earthquakes. The coseismic parameters are the maximum amplitude of coseismic water level changes, coseismic step, coseismic duration and the time when coseismic oscillation reaches its maximum amplitude. In general, the correlations are poor, and the correlation tendencies are not consistent with the commonly known tendency. This is probably due to the small differences of well-epicenter distances of these earthquakes, and big differences of magnitudes of the main shock and the aftershocks.
Figure 5 shows correlations between focal depth and coseismic parameters at the Tangshan well. Even though the correlation is not strong, the changing tendencies of the fitted curves are consistent. It is very clear that the coseismic maximum amplitude, coseismic step change, and the duration decrease as the depth increases. It also shows that when the focal depth is larger than 15 km, the time when coseismic oscillation reaches its maximum amplitude is close to the surface wave arrival time. For those reaching their maximum amplitudes between the P and S wave arrivals, the focal depths are always shallower than 15 km.
Figure 6 shows that the correlations between maximum amplitude of coseismic water level changes and coseismic step change, and between maximum amplitude and duration are larger than 0.9, indicating that they are strongly correlated. There exists relatively weaker correlation between maximum amplitude and the time when coseismic oscillation reaches its maximum amplitude.
The Wenchuan MS8.0 and its 15 larger aftershocks caused prominent coseismic water level changes at the Tangshan well. From the relationship between the magnitude and well-epicenter distance, we can conclude that the Tangshan well is very sensitive to detect coseismic changes. The correlations between coseismic parameters (coseismic maximum amplitude, coseismic step, coseismic duration) and earthquake statistics (magnitude, well-epicenter distance and focal depth) shows that there are strong correlations between the coseismic parameters and earthquake magnitude, whereas the correlations between the coseismic parameters and the other two earthquake statistics, the focal depth and well-epicenter distance, are relatively weaker. This is because that the Wenchuan earthquakes are driven from the same source area, and that the well-epicenter distance and focal depth have relatively smaller differences. Therefore, magnitude is the main factor affecting the coseismic responses. Furthermore, comparing the relationship of MW, MS, and ML magnitudes versus coseismic water level responses, it shows that the correlation of MS magnitude is the strongest, whereas that of ML magnitude is the weakest. This indicates that the surface wave magnitude has more effect on the coseismic changes caused by medium to tele-seismics. The maximum amplitude of coseismic water level changes has very strong correlation with coseismic step and duration. The three coseismic variables may have been affected by dynamic strain produced by the propagation of seismic waves, and may have little relation with the static strain deriving from focal fault fracture.
Analysis on the temporal process of coseismic water level changes shows ① the Wenchuan MS8.0 earthquake causes oscillation lasting for four hours. After comparing it with the seismogram at Douhe seismic station, we get the same cognition as that of Woodcook and Roeloffs (1996) and Shu et al (2006), i.e., the well-aquifer system is equivalent to a broad-band seismometer. It also shows that not only the Rayleigh wave can cause coseismic water level, P and S wave can also cause coseismic changes (Zhang et al, 1998); ② Comparing the time when coseismic oscillation reaches its maximum amplitude with seismic phase arrivals of the Wenchuan aftershocks, there are maximum amplitudes occurring after the P wave arrivals and near the surface wave arrivals. The maximum amplitudes of the coseismic water level changes of EQ#5、EQ#6 are very close to the S wave arrivals. There are seven earthquakes (#3、#4、#7、#10、#11、#12) the coseismic changes caused by which have maximum amplitudes between S and Love wave arrivals. The coseismic changes of six earthquakes (#1、#8、#9、#13、#15、#16) reach their maximum amplitudes near the Rayleigh wave arrivals. Considering the time error of analog water level, which is about 2 minutes, it is reasonable that the maximum amplitudes of coseismic water level changes are distributed between S and Rayleigh wave arrivals. This suggests that the Rayleigh wave is not the only factor causing the maximum amplitude of water level. On the one hand the coseismic water level changes can be affected by the features of well-aquifer system for some characteristic period of which may will magnify the oscillation amplitude of water level changes, (Cooper et al, 1965; Kunugi et al, 2000); on the other hand, coseismic changes may relate to the environmental state of the borehole when seismic waves arrive. The dynamic strain caused by seismic waves at the borehole is related to the features of seismic waves and their propagation, seismic phases coupling, and hydro-geological settings around the borehole. It is the coeffect of these elements.
As for coseismic step of water level, because the well-epicenter distances of Wenchuan earthquakes are between 1 350 km and 1 650 km, and because the earthquakes are driven from the same earthquake-gestating fault, the static strain produced by the falut cracking is tiny at those distances, and can not be used to interpret the coseismic step. The strong correlations between the amount of coseismic step and earthquake magnitude, maximum amplitude and duration may suggest that those changes have the same force source, which is derived from seismic waves. The relationship between the step direction of coseismic water level changes and focal mechanism indicates that there is no obvious correlation. However, it shows consistency between step direction and water level changing tendency (Table 1). This indicates that coseismic step has stronger correlation with the propagation of seismic waves and the hydro-geological environment around the borehole. This result is similar to that of some other researchers. Koizumi et al (2004) suggested that using focal mechanism to explain coseismic changes has big discrepancy with the fact. Wang et al (2005) suggested that the causation of coseismic variation is the movement of ground surface and its aftereffect caused by seismic waves. Lai et al (2006) concluded that coseismic steps are affected by dynamic strain and hydro-geological environments. If the time serving system and the sampling rate of digital water level can be improved in the future (Shu et al, 2006), using broad-band seismograph, strainmeter and considering the hydrological environment around borehole simultaneously can make further progress in the mechanism study of coseismic water level changes.
This study is supported by National Natural Science Foundation of China (No. 40574020) and Basic Research item of Institute of Earthquake Science, China Earthquake Administration (No. 0207690236). The authors are grateful to the staff members at Tangshan Well who supplied the high standard data, and pay great thanks to the constructive comments and suggestions from the anonymous reviewers.
|
Cao X L (1988). Analysis on the tendencious rising of water level in Tangshan area. North China Earthquake Sciences 6(3): 89-92 (in Chinese with English abstract).
|
|
Che Y T and Yu J Z (2004). The annual anomaly caused by rain in Tangshan area. In: Investigation and Study on Typical Abnormity of Subsurface fluid. Meteorological Press, Beiiing, 3-10 (in Chinese).
|
|
Cooper H H, Brredehoeft J D, Papadopulous I S and Bennett R R (1965). The response of well-aquifer systems to seismic waves. J Geophys Res 70(16):3 915-3 926. doi: 10.1029/JZ070i016p03915
|
|
Hu X P, Yu C Q, Tao K, Cui X F, Ning J Y and Wang Y H (2008). Focal mechanism solutions of Wenchuan earthquake and its strong aftershocks obtained from initial P wave polarity. Chinese J Geophys 51(6): 1 711-1 718 (in Chinese with English abstract). http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=dqwlxb200806011
|
|
Huang F Q, Chi G C, Xu G M, Jian C L and Deng Z H (2000). Research on the response anomalies of subsurface fluid in mainland monitoring network to the Nantou earthquake with MS=7.6. Earthquake 20(Supp1.): 119-125 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTotal-DIZN2000S1019.htm
|
|
King C Y, Azuma S, Igarashi G, Ohno M, Saito H and Wakita H (1999). Earthquake-related water-level changes at 16 closely clustered wells in Tono, central Japan. J Geophys Res, 104(B6): 13 073-13 082. doi: 10.1029/1999JB900080
|
|
Kitagawa G and Matsumoto N (1996). Detection of coseismic changes of underground water level. Journal of the American Statistical Association 91(434): 521-528. doi: 10.1080/01621459.1996.10476917
|
|
Kitagawa Y, Koizumi N, Takahashi M, Matsumoto N and Sato T (2006). Changes in groundwater levels or pressures associated with the 2004 earthquake off the west coast of northern Sumatra (M9.0). Earth Planets Space 58: 173-179. doi: 10.1186/BF03353375
|
|
Koizumi N Y, Kitagawa M, Takahashi M, Sato T, Matsumoto N, Ito H, Kuwahara Y and Cho A (2002). Changes in groundwater level and crustal strain in and around the Kinki district related to the 2001 Geiyo earthquake. Zishin 55: 119-127.
|
|
Kunugi T, Fukao Y and Ohno M (2000). Underdamped responses of a well to nearby swarm earthquakes off the coast of Ito City, central Japan, 1995. J Geophys Res 105: 7 805-7 818. doi: 10.1029/2000JB900020
|
|
Lai W J, Koizumi N, Matsumoto N, Kitagawa Y, Lin C W, Shieh C L and Lee Y P (2004). Effects of seismic ground motion and geological setting on the coseismic groundwater level changes caused by the 1999 Chi Chi earthquake, Taiwan. Earth Planets Space 56: 873-880. doi: 10.1186/BF03352534
|
|
Liu Y W, Yang X H, Liu Y M and Chen D Q (2005). Response characters of subsurface fluid to the MS=9. 0 earthquake in Sumatra. In: Department of Earthquake Monitoring and Prediction, China Earthquake Administration ed. The 2004 MS=8. 7 Earthquake in Sumatra, Indonesia and its Influence on China Mainland. Seismological Press, Beijing, 131-258 (in Chinese).
|
|
Ma L, Yin B J, Huang J P, Chen H Z and Zhang C J (2008). Characteristics of coseismic temperature changes at Tangshan well. In: Ren Jinwei, Yin Chaomin, Earthquake Prediction Research in China. Seismological Press, Beijing, 356-363 (in Chinese).
|
|
Matsumoto N (1992). Regression analysis for anomalous changes of ground water level due to earthquake. Geophys Res Lett 19(12): 1 193-1 196. doi: 10.1029/92GL01042
|
|
Shi Y L, Cao J L, Ma L and Yin B J (2007). Tele-seismic coseismic well temperature changes and their interpretation. Acta Seismologica Sinica 20(3): 280-289. doi: 10.1007/s11589-007-0280-z
|
|
Shu Y L, Zhang S M and Huang F Q (2006). A study on earthquake-time of Zhouzhi deep borehole on two strong earthquakes magnitude 8.7 and magnitude 8.5 in Indonesia. Seismological and Geomagnetic Observation and Research 27(2): 16-22 (in Chinese with English abstract). http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=dzdcgcyyj200602004
|
|
Wang C H, Wang C Y, Kuo C H and Chen W F (2005). Some isotopic and hydrological changes associated with the 1999 Chi-Chi earthquake, Taiwan. Island Arc 27:37-54. doi: 10.1111-j.1440-1738.2004.00456.x/
|
|
Wang W M, Zhao L F, Li J and Yao Z X (2008). Rupture process of the MS8.0 Wenchuan earthquake of Sichuan, China. Chinese J Geophys 51(5): 1 403-1 410 (in Chinese with English abstract). http://en.cnki.com.cn/Article_en/CJFDTOTAL-DQWX200805014.htm
|
|
Woodcock D and Roeloffss E A (1996). Seismically induced water-level oscillation in a fractured-rock aquifer well near Grants Pass, Oregon. Oregon Geology 58: 27-33. http://ci.nii.ac.jp/naid/20001125756
|
|
Yang Z Z, Deng Z H, Zhao Y X and Zhu P Y (2005). Preliminary study on coseismic steps of water-level in Dazhai well, Simao city, Yunnan Province. Acta Seismologica Sinica 18(5): 611-617. doi: 10.1007/s11589-005-0040-x
|
|
Zhang G M, Zhang X D, Liu J, Liu Y W, Tia Q J, Hao P, Ma H S and Jiao M R (2005a). The effects of Sumatra earthquake with magnitude 8.7 in China's mainland. Earthquake 25(4): 15-25 (in Chinese with English abstract). http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=diz200504002
|
|
Zhang S X, Zhang Z G, Zhanng Y G, Sun P Q and Luo Y X (2005b). Study on the earthquake reflecting ability of digital water level and geothermals data with Indonesia strong earthquake. North China Earthquake Sciences 23(3): 27-30 (in Chinese with English abstract). http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=hbdzkx200503005
|
|
Zhang S X, Yang W D and Zhang Z G (2007). Comparison of the ability of recording seismic wave between digital and anologue groundwater level observation in Tangshan mine well. Northwestern Seismological Journal 29(2): 170-173 (in Chinese with English abstract). http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=xbdzxb200702015
|
|
Zhang Z G, Wan D K and Dong S Y (1998). Research and application of comparison between water level vibration and face wave. Earthquake 18(4): 399-404 (in Chinese with English abstract). http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=QK199800042142
|
|
Zhang Z G, Zhang S X, Zheng Y Z and Wang H S (2002). Analysis on the ability of groundwater level for recording seismic wave in Shanxi water 2 and Yue 42 wells, Tangshan, China. Northwestern Seismological Journal 24(3): 263-266 (in Chinese with English abstract). http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=xbdzxb200203013
|
Figures(6) / Tables(1)
震球期刊
EQS
Supported by: Beijing Renhe Information Technology Co., Ltd. 
|
|
DownLoad:
CSV
