Co-seismic response of underground fluid can effectively and directly reveal the stress-strain response of the crust media. Over a few decades, an earthquake can cause a wide range co-seismic response of water level and water temperature with sharp amplitude changes, which is a focus of common concern of Chinese seismologists (Huang Fuqiong et al., 2000; Liu Yaowei et al., 2005; Yan Rui et al., 2009; Yang Zhuzhuan et al., 2005, 2007, 2008; Gao Xiaoqi et al., 2006; Liao Lixia et al., 2009; Gu Shenyi et al., 2010). These studies not only focused the morphology of co-seismic response, but tried to explore the application of co-seismic response characteristics to earthquake prediction practice. Research has shown that, when the well water level rises generally over a tectonic zone, it means enhancement of compression and decrease of extension of the tectonic stress field, and water level drop represents the weakening of compression and enhancement of extension of the tectonic stress field (Huang Fuqiong et al., 2000). Therefore, the correlation between the spatial distribution of the observation wells observing the step changes of water level and the tectonic region is of positive significance in determining the change of tectonic stress field. Co-seismic responses caused by seismic waves are mainly in the forms of fluctuation in water level and drop-restoration in water temperature. The range of water level fluctuation and water temperature decrease has a certain quantitative relationship with the epicentral distance and earthquake magnitude, the mechanism of which is complex. Currently, there are three assumptions widely used to explain them, i.e. gas escape, thermal dispersion and cold water infiltration. Each of the views is reasonable and can explain the observed phenomena to a certain extent, but which mechanism is more universal and rational remains to be further studied.
On April 25, 2015, a magnitude 8.1 earthquake happened in Nepal. The earthquake caused extensive co-seismic response of underground fluid in the Yunnan region in well water level, water temperature and gas radon, water quality (calcium ion, bicarbonate ion, and PH value) in some wells, and there are two cases of macroscopic phenomena observed.
This paper systematically collected and analyzed the underground fluid co-seismic response in Yunnan caused by this earthquake, analyzed and discussed the temporal-spatial distribution, the influencing factors and the correlation of seismic responses to the digital water level and water temperature observations.1 OVERVIEW OF UNDERGROUND FLUID OBSERVATION IN THE YUNNAN AREA
Through the updating and digital transformation of underground fluid observation instruments and equipment in the Yunnan region in the last few decades, great progress has been made in underground fluid observation of the Yunnan region. So far, the underground fluid level, water temperature and water chemistry observation networks in the Yunnan region contain a total of 300 observation items (excluding auxiliary measuring items), namely, water level observation items: 70, water temperature 73, radon 54, mercury 32, water quality (ions and others) 71, a total of 300 items. Statistics of the observation items are shown in Table 1.
For digital water level and water temperature observations, the sampling rate is one per minute, and gas radon and mercury is sampled per hour. For water quality including ions and PH, and other observation items, they are analog observations and measured once a day. Among them, the only digital observation is helium, and the sampling rate is one per hour.2 CO-SEISMIC RESPONSE OF UNDERGROUND FLUID TO THE MS8.1 NEPAL EARTHQUAKE
The epicenter of the April 25, 2015, Nepal MS8.1 earthquake was 1, 430km and 1, 960km away from the nearest and farthest well of the underground fluid observation well network in Yunnan. After the earthquake, co-seismic response was observed in 39 wells of the 45 digital water level observation wells in the Yunnan regional underground fluid observation well network; co-seismic reponse was observed in 16 of the 48 digital temperature observation wells, of the 20 digital gas-radon observation wells, one well at Qujiang recorded the co-seismic response, among the 15 analog water quality observation wells, the well in Pu'er observed co-seismic response in the measurements of calcium ion, bicarbonate ion, and the value of PH. Additionally, two macroscopic phenomena where water turned muddy were found in Lijiang and Binchuan. The distribution of wells with co-seismic response is shown in Fig. 1. Observation items that responded to this earthquake are shown in Fig. 2, and there was no co-seismic response observed in gas-mercury.
From Figs. 1 and 2 it can be clearly seen that water level and water temperature observation wells with co-seismic response were not only large in quantity but also widespread, thus having a significantly higher earthquake recording ability for far-field earthquakes than gas radon and water quality observations. Although the data shows that individual gas-radon stations can record far-field earthquakes, its identifiability is significantly lower than that of the water level and water temperature observations.3 THE SEISMIC RESPONSE CHARACTERISTICS OF THE DIGITAL WATER LEVEL AND WATER TEMPERATURE OBSERVATIONS TO THE NEPAL MS8.1 EARTHQUAKE
Analysis on original images of the data inputs of digital water level meter and water temperature metershowed that co-seismic variations related to the May 24, 2015 MS8.1 Nepal earthquake were recorded by a total of 45 sets of equipment in 39 water level observation wells and 20 sets of equipment in 16 water temperature observation wells in the Yunnan area. Characteristics of the records are summarized as follows:
(1) Among 45 normal digital water level observation wells, 39 wells observed the co-seismic response, accounting for 78% of the observation wells. In 48 digital water temperature observation wells, 16 wells observed seismic response, accounting for 33% of the observation wells. It can be seen from the co-seismic response distribution of water level and water temperature in Fig. 3 that, the variation forms of co-seismic response of water level and water temperature varied to a certain extent. The co-seismic response of water level appeared in the forms of step rise, step drop and fluctuation, while the co-seismic response of water temperature showed a pattern of rise or fall, then recovery.
(2) As can be seen from the statistics of water level changes in Table 2, of the 39 well water level observation items, 15 observations are of step change (step rise: 11, and step drop: 4), with the step-rise as the dominant change, and the rest 24 of the observation items are of fluctuations. The maximum amplitude and duration of co-seismic change of the MS8.1 Nepal earthquake recorded by different wells varied greatly, for example the amplitude of change is 5 mm at Lanping, 940mm at Nujiang, and the duration is 4 minutes at Lanping and 540 minutes at Qujing; the amplitude of fluctuation is 1mm at Gengma and 1, 096mm at Zhaotong, the duration is 2 minutes at Gengma and 74 minutes at Kunming station (Fig. 4).
Statistical analysis was made on the step changes of the 15 water level measurements and 24 fluctuations measurements to obtain the relationship of amplitude of co-seismic change with duration and epicentral distance, respectively. As seen from Fig. 5, the length of duration is proportional to the maximum amplitude, the larger the maximum amplitude of water level change, the longer the duration (except a few wells), however, the amplitude variation has no statistical relationship with epicentral distance.
(3) As seen from the statistical changes of water temperature in Table 2, of the 16 well water temperature measurements, 8 are of temperature increase and 8 are of temperature decrease. The maximum amplitude and duration of co-seismic change of the MS8.1 Nepal earthquake recorded by different wells varied greatly. As shown in Fig. 6. the maximum double amplitude is 0.0006℃ at Xundian and 0.1169℃ at Zhaotong, the duration is 18 minutes at Qujiang and 305 minutes at Fuming (Fig. 6).
Statistical analysis was made with the 16 water temperature measurements to obtain the relationship of amplitude of co-seismic change with duration and epicentral distance, respectively. As seen from Fig. 7, there is no significant statistical relationship between them.
(4) After the Nepal MS8.1 earthquake, the largest aftershock occurred on May 12, 2015 with magnitude of 7.5. As seen from Fig. 8, the changes of co-seismic response to the main shock and the strong aftershock in water level and water temperature recorded by a same set of instruments present a similar pattern, the larger the earthquake magnitude, the greater the co-seismic response amplitude. With respect to water level, there are 5 wells that recorded the co-seismic response to the M7.5 aftershock, which are Zhaotong, Dehong, Jianshui, Chengjiang, and Yongsheng; and with respect to water temperature, three wells recorded the co-seismic response to the M7.5 earthquake, which are Zhaotong, Dehong and Lijiang, accounting respectively for 11% and 6% of the observation wells. Compared to the main shock, the number of wells with seismic response to the aftershock is less.
(5) During the Nepal earthquake, there were two different types of instrument installed in the 6 water level observation wells (Qujiang, Xiaoshao, Mile, Kaiyuan, Chengjiang, Tonghao and Gaoda) and 4 water temperature observation wells (Pu'er, Xiaguan, Chengjiang, Mile), conducting the parallel observation of water level and water temperature. As can be seen from Fig. 9, instruments of different types in a same well all recorded the co-seismic response synchronously, but the maximum amplitude and duration of the response were significantly different recorded by different models of instrument.
(6) During the Nepal earthquake, co-seismic response of water temperature was observed in 16 observation wells, of which, 15 wells also observed a water level co-seismic response in the same well (a co-seismic response was recorded in Xundian well in water temperature but not in water level) (Fig. 1, Table 2). This indicates that the co-seismic response of water temperature is related to the co-seismic response of water level.4 CONCLUSIONS AND FINDINGS
(1) The Nepal MS8.1 earthquake had a great impact on the Yunnan region, causing significant co-seismic response of underground fluid in both microscopic and macroscopic observations. Digital water level and water temperature observations have significantly higher capability to record this great earthquake than radon gas and water quality observations do. Analysis reveals the following reasons for this. First, restricted by observation techniques and methods, the current gas radon and water quality observations are the hourly value and daily value. With such a sampling frequency, the observations can hardly capture the instantaneous dynamic change in the crust media caused by seismic waves propagating rapidly through the well-aquifer system. Secondly, the radon gas, water quality and other chemical changes are not simply caused by disturbances of stress and strain, but controlled by many factors. Microscopic changes in radon gas and water quality caused by seismic waves cannot be recorded with current observation precision. Furthermore, there are many disturbing factors existing in radon gas and water quality observation, which will have an impact to a certain extent on the extraction of co-seismic response or abnormal changes (Sun Xiaolong et al., 2008).
(2) The water level co-seismic response ratio is 78%, and the water temperature co-seismic response ratio is 33%. This indicates that the water level and water temperature observations in Yunnan have a strong response capability to high-frequency crustal dynamic action, and the earthquake recording ability of the water level is higher than that of water temperature. The ratios of water level and water temperature observation wells that recorded the co-seismic response of the strongest aftershock are 11% and 6%, respectively, illustrating that the water level and water temperature co-seismic response are closely related to the magnitude of far-field earthquakes; the greater the magnitude of earthquake, the higher the response degree will be. This also means more wells will record the co-seismic response, which is consistent with the results by reference (Yang Zhuzhuan et al., 2005).
(3) The maximum amplitude and duration of water level and water temperature co-seismic response vary greatly with different wells. The morphology for water level is dominated mainly by oscillation and step-rise, and for water temperature, it is manifested as a slow recovery after rising or falling.
Different observation wells have different response characteristics tofar-field earthquakes, which is mainly related to the hydro-geological background of the observation wells (Yu Jinzi et al., 2012).
(4) The variation of water level and water temperature response to the main shock and the largest aftershock recorded by the same well has indicated that the greater the earthquake magnitude, the larger the amplitude of co-seismic response, but the morphology is similar.
Water level and water temperature synchronous response can be recorded by all types of instruments installed in a same well, but there is significant difference in the maximum amplitude and duration of the response by different instruments.
A single observation well has a relatively consistent record form to different earthquakes, and in addition to the observation instruments, the amplitude of response is related to a certain extentto the magnitude of far-field earthquakes (Liu Chenglong et al., 2009).
(5) The wells with water temperature co-seismic response all recorded water level co-seismic response. The co-seismic responses of well water level and water temperature are closely related. Water temperature co-seismic response is mainly induced by well water level co-seismic response, mostly by a mixture of upper and lower water masses in the wellbore, when the water level is oscillating (Yu Jinzi et al., 2012).
|Gao Xiaoqi, Chen Huajing, Yang Youling, et al. Co-seismic effect of the Dec.26, 2004 M8.7 Indonesia earthquake on the Xinjiang region[J]. Seismological and Geomagnetic Observation and Research, 2006, 27(3): 40–46.|
|Gu Shenyi, Li Zhixiong, Zhang Hui. Analysis of the characteristics of co-seismic records of five wells water level in Hainan area in the MS8.0 Wenchuan earthquake event[J]. Journal of Seismological Research, 2010, 33(1): 36–43.|
|Huang Fuqiong, Chi Gongcai, Xu Guiming. Research on the response anomalies of subsurface fluid in mainland monitoring network to the Nantou earthquake with MS7.6[J]. Earthquake, 2000, 20(Suppl): 119–125.|
|Liao Lixia, Wang Meiling, Wu Shaozu. Study on co-seismic effects of water level as observed by Fujian well network and its implication to earthquake prediction[J]. Acta Seismologica Sinica, 2009, 31(4): 432–441.|
|Liu Chenglong, Wang Guangcai, Zhang Weihua, Mei Jianchang. Co-seismic response of well water level in Three Gorges well network to MS8.0 earthquake in Wenchuan[J]. Acta Seismologica Sinica, 2009, 31(2): 188–194.|
|Liu Yaowei, Yang Xuanhui, Liu Yongmin et al. The characteristics of subsurface fluid response to the Sumatra magnitude 8. 7 earthquake [A]. In: Department of Monitoring and Prediction, China Earthquake Administration (ed). The 2004 Sumatra, Indonesia Magnitude 8. 7 Earthquake and Its Influence on the Chinese Mainland [M]. Beijing: Seismological Press, 2005. 131-258 (in Chinese).|
|Sun Xiaolong, Wang Bo. Analysis of earthquake-recording ability of subsurface fluid in Xiaojiang fault zone[J]. South China Journal of Seismology, 2008, 28(4): 18–26.|
|Yan Rui, Huang Fuqiong. A preliminary study on the co-seismic response of water level of Huanghua Well to the 5 earthquakes in Sumatra[J]. Earthquake Research in China, 2009, 25(3): 325–332.|
|Yang Zhuzhuan, Deng Zhihui, Liu Chunguo, et al. Analysis on the characteristics of co-seismic response of well water level and water temperature regime in China mainland to the Wenchuan MS8.0 earthquake[J]. Seismology and Geology, 2008, 30(4): 895–905.|
|Yang Zhuzhuan, Deng Zhihui, Tao Jingling, et al. Co-seismic effects ofwater temperature based on digital observation from Tayuan Well, Beijing[J]. Acta Seismologica Sinica, 2007, 29(2): 203–213.|
|Yang Zhuzhuan, Deng Zhihui, Zhao Yunxu, et al. Preliminary study on co-seismic steps of water-level in Dazhai Well, Simao City, Yunnan Province[J]. Acta Seismologica Sinica, 2005, 27(5): 569–574.|
|Yu Jinzi, Che Yongtai, Liu Chenglong, Li Wanming.. Co-seismic responses of water level and temperature in wells of the Jinshajiang Groundwater Observation Network to the Japan MW9.0 earthquake: characteristics and mechanism[J]. Earthquake, 2012, 32(1): 59–69.|