Characteristics and Seismogenic Background of the 2014 Earthquake Sequence around Shanxi Reservoir in Eastern China

doi:10.21203/rs.3.rs-4822045/v1

Download PDF

Research Article

Characteristics and Seismogenic Background of the 2014 Earthquake Sequence around Shanxi Reservoir in Eastern China

https://doi.org/10.21203/rs.3.rs-4822045/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The Shanxi Reservoir area is one of the most seismically active regions in Eastern China. We use the seismic waveforms, earthquake catalogs, and seismic phase observation reports to study and summarize the earthquake sequence characteristics and seismogenic background of the Shanxi Reservoir area in 2014. (1) Firstly, we relocate the earthquake sequence by the HypoDD method. The relocation results show that the dominant strike of the sequence is approximately NW-SE 129°, with a steep dip of roughly SW 84°, and the seismogenic structure contains multiple secondary surfaces (sub-faults). (2) Based on the relocation catalog, we estimate the spatial distribution of b-values along the strike of the seismogenic fault using the maximum likelihood method, and obtain the abnormal regions of stress and inhomogeneity near the fault. (3) At the same time, we also calculate the focal mechanisms for events with ML ≥ 2.5 and ML ≥ 3.0 by the HASH and CAP methods, respectively, and obtain rose statistical charts for each parameter. The results indicate that the 2014 sequence is basically of the right-lateral strike-slip type. Combined with the relocation results, we speculated that the seismogenic surface is close to the nodal plane with a strike of NW-SE 131°, a dip of SW 82°, and a rake of -173°. (4) On the basis of the focal mechanisms, we further acquire the azimuth and inclination parameters of the P and T axis for each earthquake, and get the rose statistical charts of these parameters. The P-axis and T-axis of the 2014 sequence are dominantly distributed in the NNW and SEE direction, with a gentle inclination angle, which exhibits nearly horizontal compression in the NNW direction and nearly horizontal tension in the SEE direction. (5) Finally, combined with the distribution of fault zones around the reservoir area, we infer that the 2014 sequence around the Shanxi Reservoir was the result of right-lateral strike-slip rupture of the f_11 − 3 branch of the Shuangxi-Jiaoxiyang fault. We find that the 2014 sequence began to occur in areas where multiple groups of faults converge and have favorable infiltration conditions, which are regions of high-stress concentration and significant infiltration effects. The infiltration of reservoir water into deeper areas leads to an increase in pore pressure and a decrease in the strength of structural surfaces, making the rock more susceptible to rupture. Under the influence of continuous ruptures (earthquakes) and infiltration of reservoir water, the 2014 earthquakes developed along the f_11 − 3 fault towards NW and SE and deeper directions.

reservoir seismicity

earthquake relocation

focal mechanism

stress field

water infiltration

Shanxi Reservoir

The Shanxi Reservoir is located in Zhejiang Province in eastern China (Fig. 1a), with a design maximum capacity of 1.824 billion m³and a maximum water level of 154.75m. Before the construction of the reservoir, no earthquake was recorded within 25 km nearby, and only three earthquakes of magnitude ~ 4.0 and three earthquakes of magnitude < 2.0 were recorded within 50 km. Seismic activity of this area was low in frequency and intensity. The reservoir began impounding in May 2000. Then in July 2002, an earthquake sequence with a maximum magnitude of ML 3.5 occurred (Fig. 1b). After that, earthquakes continued, and the Shanxi Reservoir became one of the most seismically active areas in Zhejiang Province. In 2006 and 2014, earthquake sequences with maximum magnitudes of ML 4.6 and ML 4.2 occurred, respectively, causing great economic losses and social impacts.

After the 2006 earthquakes, researchers have carried out many studies on earthquake relocation and velocity structure inversion (Li et al., 2007; Zhong et al., 2010; Zhu et al., 2010; Zhong et al., 2011), and focal mechanism solutions (Zhang et al., 2013). They obtained the seismogenic fault parameters and regional stress characteristics, and concluded that the 2006 events were the results of the initial rupture of fault f₁₁ and the combined action of multiple secondary faults such as f7, f9, and f14.

After the 2014 earthquakes, Ma et al. (2016) investigated the geologic characteristics of the major faults around the reservoir by the method of field geological survey. Yang et al. (2019, 2021) studied the deep electrical structure of the reservoir area by the magnetotelluric sounding method (MT). Hou et al. (2018) acquired the tectonic characteristics of Shanxi Reservoir from surface to mid-crust through the deep seismic reflection method. These studies found that the second and third branches (Fig. 1a, f_11 − 2 and f_11 − 3) of the fault f₁₁ have certain water permeability, which is the critical fault near the reservoir area. The third branch of the Shuangxi-Jiaoxiyang (Fig. 1a, f_11 − 3) fault is the seismogenic fault for the 2014 earthquakes (Ma et al., 2016; Yang et al., 2019; Hou et al., 2018).

After 2008, we constructed several monitoring stations near the reservoir (Fig. 1): Shanxi earthquake station (SX), Yunhu earthquake station (YH), Baoyang earthquake station (BY), Yangshan earthquake station (YS), Huangtan earthquake station (HT), Huanglong earthquake station (HL), Xinpu earthquake station (XP), and Taishun earthquake station (TSH). Therefore, for the 2014 earthquake sequence, the instruments can record more minor earthquakes, and the seismic waveform data and earthquake catalog are richer, which can provide more complete data for more detailed earthquake relocation and focal mechanism research. In addition, the development of relocation programs and focal mechanism programs provide more scientific calculation methods for the study of seismogenic structure and mechanical mechanisms. It is of great significance to find out the spatial and tectogenesis characteristics of seismogenic structures around Shanxi Reservoir, for the perception and risk assessment of reservoir earthquakes.

Based on the seismic waveforms and earthquake catalogs recorded by stations in Zhejiang and neighboring provinces, we study and summarize the ML 4.2 earthquake sequence near the Shanxi Reservoir in 2014. Firstly, we relocate the sequence to obtain a more precise spatial and temporal distribution of earthquakes. And we use the relocated earthquake catalog to estimate the spatial distribution of b-value near the fault, and infer the stress level and rock inhomogeneities distribution in the region. Then, we calculate the focal mechanisms of the ML ≥ 2.5 earthquakes, and obtain the attitude and movement characteristics of the seismogenic fault plane. Furthermore, we count the focal mechanism and stress axis parameters, to obtain the characteristics of the stress field in the region. Finally, combined with geological data, we infer the seismogenic fault, and discuss the cause of earthquakes around the Shanxi Reservoir. It helps to understand the seismogenic structure and mechanism of historical earthquakes around Shanxi Reservoir, and puts forward a better understanding for seismic hazard assessment of the reservoir.

2.1 Data

We collect earthquake catalogs, seismic phase reports, and waveform records of the Shanxi Reservoir area (119.75°E-120.1°E、27.6°N-27.8°N) from 2000 to 2023 recorded by stations in Zhejiang and neighboring provinces from Zhejiang Provincial Seismic Network. From the earthquake catalog, we plot the spatial distribution of earthquakes (Fig. 1a) and the magnitude-time image (Fig. 1b), which shows that the earthquake sequences in the region are distributed in clusters and groups in space and time.

There were few stations near the reservoir at the beginning of the impoundment, and it was not until 2008 that we established a more complete network of earthquake monitoring stations here. After 2008, there has been eight seismic stations within 30 km from Shanxi Reservoir area (Fig. 1a), and most earthquakes can be recorded by more than 4 stations, which can determine the focal depth and epicenter position more accurately.

2.2 Relocation method

A complete and comprehensive relocated earthquake catalog can help us understand the geometry of the seismogenic faults around the Shanxi Reservoir, and can provide a basis for studying the mechanism of earthquake sequences.

We use the HypoDD double-difference location method (Waldhauser and Ellsworth, 2000) to relocate the 2014 earthquake sequence around the Shanxi Reservoir. The HypoDD method firstly matches neighboring events into earthquake pairs, and then determines the relative position of the epicenters by using the residuals between the observed and theoretical values of the travel time difference of two earthquakes. This method eliminates the velocity perturbation on the seismic wave propagation path, which can effectively reduce the errors caused by the imprecise and inaccurate crustal velocity model, and significantly improves the accuracy of earthquake location. It has shown good results in the research of earthquake sequence relocation (Yang et al., 2022; Waldhauser et al., 2004; Hauksson and Shearer, 2005; Haukssonet al., 2012; Jiang et al., 2019).

The velocity model used for relocation refers to the study of Zhong et al. (2017) (Table 1). Since the picking accuracy of the P-wave seismic phase is higher than that of the S-wave seismic phase, the weight of the P-wave is set to 1 and the weight of the S-wave is set to 0.5. Parameters for seismic clusters are set as follows: MAXDIST (maximum distance between event pair and station) = 200 km、MAXSEP (maximum hypocentral separation between event pairs) = 3 km、MINLNK (minimum number of links required to define a neighbor) = 8、MINOBS (minimum number of links per pair saved) = 8、OBSCT (minimum number of links per event pair to form a cluster) = 6. In this study, the LSQR (conjugate gradient method) method is used to iteratively solve the equations to obtain the damped least squares solution.

Table 1

Velocity model of the Shanxi Reservoir area
Depth (km)	0	3	13	17	20	33
Velocity of P wave (km/s)	5.7	6.0	6.15	5.85	6.6	8.1

2.3 Methods for Focal Mechanism

The focal mechanisms can reflect the stress field in hypocenter region, which is an important parameter in investigating the cause of earthquakes and analyzing the seismogenic structures. The density of stations around Shanxi Reservoir is relatively high, and the distribution of stations is relatively uniform, which provides a good condition for calculating the focal mechanisms. In this study, we choose the widely used HASH (Hardebeck and Shearer, 2002, 2003) and CAP (“Cut and Paste”. Zhu and Helmberger, 1996) methods to calculate the focal mechanisms.

The HASH method uses the initial polarity of the P wave, or adds information such as amplitude to determine the attitude parameters of the seismogenic fault plane. This method is concise, does not need information such as instrument response, and has low signal-to-noise requirements. It is suitable for moderate and small earthquakes. Moreover, the method evaluates the quality of the calculation results in terms of station distribution and fault plane uncertainty, which are categorized into six grades: A-F. Considering the reliability of the calculation results, we firstly screen the waveform records within 200 km from the epicenter, with signal-to-noise ratios greater than 5.0 and clear seismic phases, and then select the events with ML ≥ 2.5 and with no less than 8 station records participating in the inversion, and finally adopt the focal mechanism results of A, B and C grades, totalling 99 groups.

In order to verify the reliability of the results obtained by the HASH method, we also carry out focal mechanism inversion by the CAP method. The CAP method divides broadband seismic wave records into two parts, body wave (Pnl) and surface wave, and then fits them separately, and obtains the optimal depth through the grid search method. This method can make comprehensive use of body wave and surface wave information, and can adjust the weights of body wave and surface wave, making full use of the information carried in seismic waves (Long et al., 2019). It is characterized by less dependence on velocity model and station distribution, and has relatively stable calculation results.

In the calculation using the CAP method, we first process the selected observation data by deducting instrumental response, rotating coordinates, and filtering, in which the filtering ranges of Pnl and surface waves are 0.05 ~ 0.2 Hz and 0.05 ~ 0.1 Hz, respectively. Then, based on the F-k method (Zhu and Rivera, 2002), we use the velocity model in Table 1 to calculate Green's function for different epicenter distances. Finally, we calculate the best-fitting coefficients of theoretical and actual waveforms in five dimensions (strike, dip, rake, magnitude, and depth) by the grid search method.

3.1 Relocation results

After ten iterations of the HypoDD program, the average changes in hypocenter location along the longitude, latitude and depth directions (DX, DY, DZ) decrease from 0.98 km, 0.75 km and 0.81 km at the beginning to 0.03 km, 0.09 km and 0.1km at the end, the iteration tend to be stable. The average travel time residual decreases from 0.08 s at the beginning of iteration to 0.01 s after iteration. After relocation, the travel time residual is significantly reduced, and the accuracy of the relative location of earthquakes is significantly improved. Finally, we obtain 4204 relocation results of the 2014 earthquake sequence.

Before relocation (Fig. 1a), the epicenter distribution of the 2014 sequence was more scattered. After relocation (Fig. 2), the distribution becomes more concentrated and its relationship to the tectonics becomes clearer. The epicenters show an NW-SE dominant direction, with an angle of about 135° to the north. The total length of the long axis of the epicenter distribution is about 15 km, so the rupture scale of this earthquake sequence is estimated to be about 15 km. The epicenters are mainly located in the central northeast part of the reservoir area, on the NW segment of the Shuangxi-Jiaoxiyang (f_11 − 3) fault. The earthquake distribution along the Shuangxi-Jiaoxiyang fault is segmented (Fig. 2). In order to have a clear understanding of their vertical depth distribution characteristics, we divide the area into five regions, from R1 to R5, and project them separately (Fig. 2c, Fig. 3).

In terms of the occurrence time (Fig. 2a), earthquakes start in the R2/3 regions, and then gradually spread to the R1/4/5 regions on both sides and deeper places. Earthquakes begin to occur in the area where multiple faults (for example, f₅ and f₁₁, etc.) converge, in the area with dense structural planes and good permeability. The geometric complexity of faults may cause localized stress concentration, leading to the initial rupture of this earthquake sequence. Then, the earthquakes rupture from the initial location to both sides along the SE and NW directions, and the sequence gradually terminates at both ends of the dominant spreading direction. The NW spread terminates at the intersection of f₁₁ and f₃ faults, which may be related to the change of fault slip and the blockage of reservoir water infiltration (Yang et al., 2021).

Figure 2b and Fig. 3 show that the average hypocenter depth is 4 km, the minimum depth is less than 2 km, and the maximum depth is more than 6 km. Most earthquakes are shallow, and even though their magnitude is small, they are easily perceived and prone to cause serious impacts and injuries. The profile projection perpendicular to the dominant direction (Fig. 2c, profile AA’) of epicenters can reflect the dip characteristics of seismogenic structure (Fig. 3). It is roughly estimated that the dominant dip angles in the R1-3 region are about 84° (Figs. 3a-c, f-h); the seismogenic faults are characterized by high dip angle and dipping to SW. Especially in the R4 region, there are several secondary faults dipping to NE with a high-dipping angle of 86°, and the structure of the seismogenic faults is relatively complex. Therefore, we speculate that the f₁₁ fault is relatively broken, and the secondary structures are most developed near the R4 region.

The earthquake magnitude (Figs. 2c and 3) combined with the occurrence time (Fig. 2a) shows that the earthquakes are strong at the beginning of the sequence. With the occurrence of the sequence, the magnitude gradually decreases and the earthquake location extends towards both NW and SE ends. We also calculate the spatial distribution of b-values (Fig. 2d) based on the magnitude distribution in Fig. 2c.

In the magnitude-frequency relation log₁₀N = a-b*M proposed by Gutenberg and Richter (1944), the value of b describes the proportionality of large and small earthquakes in the sequence. Rock mechanics experiments, statistical analysis, and numerical simulations found that the b-value is usually negatively correlated with regional stress (e.g., Schorlemmer et al., 2005; Kun et al., 2013; Main et al., 1992; Nuannin et al., 2012; Scholz, 2015; Wiemer and Benoit, 1996). That is, the stress accumulation is more significant in areas with lower b-values. Mogi (1962) considered the inhomogeneity of the underground medium as an important factor influencing the value of b, and the higher the degree of inhomogeneity, the larger the b-value. Figure 2d shows that the b-values are lower at the location where the earthquake sequence started (black dashed circle in Fig. 2d). We hypothesize that this is because the faults intersect at the earthquake starting location, and the stress is more likely to concentrate in the area where structural surfaces intersect. Furthermore, the white dashed line in Fig. 2d circles the area where high and low b-values coexist. The high b-values are in the area where the water body is located. We hypothesize that the infiltration of the reservoir water led to higher water content in the rock below, i.e., higher degree of heterogeneity in the underground medium. The low b-value areas are located on land, indicating that the stress is relatively concentrated here.

According to the results of relocation, it can be concluded that the strike of the seismogenic fault is NW-SE, with an angle of 129° to north, a higher dip of about 84°, dipping to SW. The earthquake sequence began in the R2 region and progressed to both the NW and SE sides. The seismogenic fault structure is relatively complex, and there may be multiple sub-faults.

3.2 Focal mechanism results

The ML ≥ 3 focal mechanisms calculated by the HASH and CAP methods are shown in Fig. 4a and c. The azimuth and inclination of the P and T axis (Fig. 4b and d) can be calculated using the focal mechanism parameters. The consistency of both the focal mechanism and the P/T-axis azimuth for the 2014 sequence around Shanxi Reservoir is high (Fig. 4). It is inferred that the seismogenic characteristics of these events are similar, and there is no obvious stress rotation during the occurrence of earthquake sequence. In order to have a clearer understanding of the overall characteristics of the above parameters, we count the frequencies of the strike, dip, and rake for each nodal plane, and the azimuth and inclination for the P/T axis (Fig. 5).

From Figs. 4a, c and 5a, the plane I is dominant in NW-SE direction, and the plane II is dominant in NE-SW. The dip angles of nodal planes I and II are concentrated in the range of 70°-90°, so the dip angle of seismogenic surface is relatively steep. The rake angle is close to horizontal, and the fault surface movement is characterized by strike-slip, which indicates that the slip of seismogenic faults is mainly horizontal dislocation. Combined with the dominant NW-SE strike of the earthquake sequence (Figs. 2 and 3), it can be inferred that the nodal plane I is the main rupture surface of the earthquakes, with a SW dipping, and the strike of the fault surface is close to that of Shuangxi-Jiaoxiyang fault (f₁₁, Fig. 1). Then, in Figs. 4b, 4d, 5b and 5d, the compressive P-axis is dominantly distributed in the NNW, near north direction; and the tensile T-axis is oriented in the NEE, near east direction. The inclination angles of the P and T axes are mostly less than 20°, which are nearly horizontal. The stresses near the fault show a sub-horizontal north-south compression, and a sub-horizontal east-west tension, resulting in a right-lateral strike-slip movement of the seismogenic fault.

To verify the reliability of the statistical results, we use the method proposed by Wan (2019) to obtain the center solution of focal mechanism for earthquakes with magnitude ML ≥ 3: the strike, dip and rake of nodal plane I are 131°, 82° and − 173°, respectively; and the strike, dip and rake of nodal plane II are 40°, 83° and − 8° respectively. The P-axis azimuth of the central solution is NNW 355° with an inclination of 11°, and the T-axis azimuth is NEE 85° with an inclination of 1°. Above results are consistent with the statistical rose chart (Fig. 5).

Based on the focal mechanisms and stress axis results, we infer that the 2014 earthquake sequence around Shanxi Reservoir was a right-lateral strike-slip activity with a strike NW-SE, dipping SW, and high dip angel, under the stress conditions of gentle-inclination N-S compression and W-E tension.

4.1 The seismogenic structure

The relocation results show that the seismogenic fault plane strikes NW 129°, with a dip of SW 84° (Figs. 2 and 3). It is close to the parameters of one nodal plane (strike NW-SE 131°, dip SW 82°, rake − 173°) of the center focal mechanism solution (Figs. 4 and 5). It is basically consistent with the location and attitude of the third branch of the Shuangxi-Jiaoxiyang fault (Fig. 1, f_11 − 3), which were obtained from the field geological survey (Ma et al., 2016), the magnetotelluric sounding method (Hou et al., 2018), and the deep seismic reflection method (Yang et al., 2019, 2021). In addition, the projection of the relocated epicenters on the AA' profile (Fig. 3) shows that the seismogenic fault is relatively complex, which is consistent with the phenomenon that secondary surfaces, fractured zones, and dikes developed on the f_11 − 3 fault, as discovered by Ma et al. (2016) through the field geological survey. Therefore, we conclude that the 2014 earthquake sequence in the Shanxi Reservoir area was the result of right-lateral strike-slip ruptures of the NW-strike Shuangxi-Jiaoxiyang fault.

Our focal mechanism results are close to those of previous studies on the 2006 earthquake sequence around Shanxi Reservoir. We speculate that both earthquake sequences may be the result of the right-lateral strike-slip rupture of the Shuangxi-Jiaoxiyang fault (f_11 − 3), and there is a phenomenon of seismic activity migrating westward along the f_11 − 3 fault.

4.2 Causes of earthquakes

The effects of reservoir water on the surrounding area include: elastic response caused by water loading (or unloading), pore pressure changes caused by diffusion or elastic loading/unloading, and the weakening of rock by the physical and chemical action of reservoir water (Gupta, 2002; Gough and Gough, 1970ab; Talwani, 1997).

The 2002, 2005, and 2006 earthquake sequences occurred during periods when the reservoir water level dropped sharply or changed frequently (Fig. 1b). It is hypothesized that the loading and unloading of water has a significant impact on seismicity at that time period. However, before the 2014 sequence, the water level was relatively stable, and there may be little correlation between the seismic activity and the loading (or unloading) of reservoir water during this period.

The 2002 sequence mainly occurred on the middle segment of the f₁₁ fault, the 2006 sequence migrated towards the southeast segment of the fault, and the 2014 sequence was mainly distributed on the northwest segment of the f₁₁ fault (Fig. 1a). This migration process may reflect the infiltration process of reservoir water and the adjustment of stresses. In other words, after the reservoir was impounded in 2000, on the permeable f_11 − 2 fault (Fig. 1a), the reservoir water infiltrated deeper along the fault face and fissures. Earthquakes around the reservoir began to be active. The occurrence of the 2002 earthquakes could create new infiltration channels, allowing reservoir water to infiltrate into the rock surrounding the f_11 − 3 fault. Then, under the tectonic stress field, the f_11 − 3 fault experienced right-lateral strike-slip dislocation, leading to the earthquake activity in 2006. Afterwards, activity on the f_11 − 3 fault further intensified and migrated westward, and the 2014 earthquake sequence occurred on the NW segment of the fault.

The initial rupture of the 2014 sequence around Shanxi Reservoir occurred at the fault intersection (the red circle Fig. 1; Fig. 2), and developed along the f_11 − 3 fault to its both sides and at greater depths (Figs. 2 and 3), concentrating within the block enclosed by f₃,₁₁,₁₃,₁₄ faults. Studies (Ma et al., 2016; Zhong et al., 2017; Yang et al., 2019, 2021) have shown that the second and third branches (Fig. 1a, f_11 − 2/f_11 − 3) of the Shuangxi-Jiaoxiyang fault have certain degree of water permeability. In the area circled in Fig. 1, there are a series of high-dip tensile structural fissures with large longitudinal scales and good connectivity, which can easily cause reservoir water to infiltrate deeper. However, other fault zones in the reservoir area are more cemented and have poor water infiltration conditions.

The faults in the Shanxi Reservoir area intersect with each other, forming fissures or cross-sectional structures with certain geometry and physical characteristics within underground rock. When the reservoir is impounded, these structures provide channels for infiltration of reservoir water. Reservoir water infiltrates into deeper, resulting in the increase of pore pressure (caused by diffusion) and the decrease of structural surface strength (caused by the physical and chemical action of reservoir water), which makes rock more prone to dislocation. Moreover, the intersections of structural surfaces in underground rock are usually stress concentration zones, which are areas where the rock is susceptible to failure. Therefore, the 2014 earthquakes began to occur in areas where multiple faults intersect and infiltration conditions are favorable (the red circle in Fig. 1), and distributed near fault zones where infiltration has a strong influence. Under the influence of seismic activity (the occurrence of earthquakes and the continued rupture of the rock create new infiltration channels) and continuous infiltration of reservoir water, earthquakes developed along the f_11 − 3 fault towards both the NW and SE ends and the deeper direction.

4.3 Reliability of results

The attitude (strike, dip, rake) characteristics of the seismogenic surface revealed by the relocation results and the focal mechanism results are in good agreement. And the results obtained using different focal mechanism methods (CAP and HASH) are close to each other (Fig. 4). Moreover, they are consistent with the geometry of the seismogenic faults obtained by the field geological survey method (Ma et al., 2016), the magnetotelluric sounding method (MT) (Hou et al., 2018), and the deep seismic reflection method (Yang et al., 2019; 2021). All of these have verified the reliability of our calculation results.

The average optimal epicenter depth of earthquakes with magnitude ML ≥ 3 calculated by the CAP method is about 4 km, which is close to the average depth of the same earthquakes obtained by the HypoDD. The results of the two methods indicate that the events around Shanxi Reservoir are shallow earthquakes, demonstrating the reliability of the inversion results in our study. On the other hand, the HASH can evaluate the quality (Grades A-F) of the inversion results in terms of parameters such as station distribution, initial motions, and amplitude ratios (Hardebeck and Shearer, 2002, 2003). The quality of the results selected in this paper is above C, which is often considered to be relatively reliable (Guo et al., 2017; Meng et al., 2022; Pei et al., 2022).

Based on the seismic waveform records, earthquake catalogs and earthquake observation reports, we analyze the characteristics of the 2014 earthquake sequence around Shanxi Reservoir. We relocate the sequence using the HypoDD method, calculate the focal mechanisms for the sequence by the CAP and HASH methods, and count the frequency of parameters of focal mechanisms and P/T axis, further discuss the seismogenic structural and tectonic stress characteristics of the earthquakes, and analyze the causes of the Shanxi Reservoir earthquakes. The main conclusions are as follows:

(1) The relocation results show that the dominant strike of the 2014 sequence is approximately NW-SE 129°, with a steep dip of roughly SW 84°, and the seismogenic structure contains multiple secondary surfaces (sub-faults).

(2) The results of the focal mechanism indicate that the 2014 sequence is basically of the right-lateral strike-slip type. Combined with the relocation results, we speculate that the seismogenic surface is close to the nodal plane with a strike of NW-SE 131°, a dip of SW 82°, and a rake of -173°.

(3) The P and T axis of the 2014 sequence are dominantly distributed in the NNW direction and SEE direction, with a gentle inclination, which exhibit nearly horizontal compression in the NNW direction and nearly horizontal tension in the SEE direction. Under such stress field conditions, the seismogenic fault undergoes right-lateral strike-slip dislocations.

(4) Combined with the distribution of fault zones around the reservoir area, we infer that the 2014 sequence around the Shanxi Reservoir was the result of right-lateral strike-slip rupture of the f_11 − 3 branch of the Shuangxi-Jiaoxiyang fault.

(5) The 2014 sequence began to occur in areas where multiple groups of faults converge and have favorable infiltration conditions, which are regions of high-stress concentration and significant infiltration effects. The infiltration of reservoir water into deeper areas leads to an increase in pore pressure and a decrease in the strength of structural surfaces, making the rock more susceptible to rupture. Under the influence of continuous rupture and infiltration of reservoir water, earthquakes developed along the f_11 − 3 fault towards NW and SE and deeper directions.

The authors have no relevant financial or non-financial interests to disclose.

Author Contribution

Yajing Gao, Qi Zhang, Yunqiang Sun and Yaqi Gao wrote the main manuscript text, and Yajing Gao prepared figures 1-5. All authors reviewed the manuscript.

Gough D. I., and Gough W. I. (1970a). Stress and deflection in the lithosphere near lake kariba–i. Geophysical Journal of the Royal Astronomical Society.
Gough, D. I. and Gough, W. I. (1970b). Load-induced earthquakes at lake kariba-ii. Geophysical Journal of the Royal Astronomical Society, 21(1), 79-101.
Gutenberg, B., & Richter, C. F. (1944). Frequency of earthquakes in California. Bulletin of the Seismological Society of America, 34(4), 185-188. doi:10.1038/156371a0
Gupta H. K. (2002). A review of recent studies of triggered earthquakes by artificial water reservoirs with special emphasis on earthquakes in koyna, india. Earth-Science Reviews, 58(3-4), 279-310.
Hardebeck, J. L., & Shearer, P. M. (2002). A new method for determining first motion focal mechanisms. Bulletin of the Seismological Society of America, 92, 2264-2276.
Hardebeck, J. L., & Shearer, P. M. (2003). Using S/P Amplitude Ratios to Constrain the Focal Mechanisms of Small Earthquakes. Bulletin of the Seismological Society of America, 93, 2434-2444.
Hauksson, E., & Shearer, P. (2005). Southern California hypocenter relocation with waveform cross-correlation, Part l: results using the double difference method. Bulletin of the Seismological Society of America, 95(3), 896-903.
Hauksson, E., Yang, W., & Shearer, P. M. (2012). Waveform relocated earthquake catalog for southern California (1981 to June 2011). Bulletin of the Seismological Society of America, 102(5), 2239-2244.
Jiang, J. Z., Li, J., & Fu, H. (2019). Seismicity analysis of the 2016 Ms 5.0 Yunlong earthquake, Yunnan, China and its tectonic implications. Pure and Applied Geophysics, 176(3), 1225-1241.
Kun, F., Varga, I., Lennartz-Sassinek, S., & Main, I. G. (2013). Approach to failure in porous granular materials under compression. Physical Review E, 88(6), 62207. doi:10.1103/physreve.88.062207
Long, F., Yi, G. X., Wang, S. W., Qi, Y. P., Zhao, M., & Agency, S. E. (2019). Geometry and tectonic deformation of the seismogenic structure for the 8 August 2017 MS 7.0 Jiuzhaigou earthquake sequence, northern Sichuan, China. Earth and Planetary Physics, 3(3), 253-267.
Main, I. G., Meredith, P. G., & Sammonds, P. R. (1992). Temporal variations in seismic event rate and b values from stress corrosion constitutive laws. Tectonophysics, 211, 233-246.
Mogi, K. (1962). Study of elastic shocks caused by the fracture of heterogeneous materials and its relation to earthquake phenomena. Bulletin of Earthquake Research Institute, 40(1), 125-173.
Nuannin, P., Kulhánek, O., & Persson, L. (2012). Variations of b values preceding large earthquakes in the Andaman-Sumatra subduction zone. Journal of Asian Earth Sciences, 61, 237-242.
Scholz, C. H. (2015). On the stress dependence of the earthquake b value. Geophysical Research Letters, 42(5), 1399-1402. doi:10.1002/2014GL062863
Schorlemmer, D., Wiemer, S., & Wyss, M. (2005). Variations in earthquake-size distribution across different stress regimes. Nature, 437(22), 539-542. doi:10.1038/nature04094
Talwani P. (1997). On the nature of reservoir-induced seismicity. Pure & Applied Geophysics, 150(3-4), 473-492.
Waldhauser, F., & Ellsworth, W. L. (2000). A double-difference earthquake location algorithm: Method and application to the northern Hayward fault. Bulletin of the Seismological Society of America, 90, 1353-1368.
Waldhauser, F., Schaff, D., Richards, P., & Kim, W.-Y. (2004). Lop Nor revisited: Underground nuclear explosion locations, 1976-1996, from double difference analysis of regional and teleseismic data. Bulletin of the Seismological Society of America, 94(5), 1879-1889.
Wiemer, S., & Benoit, J. P. (1996). Mapping the b-value anomaly at 100 km depth in the Alaska and New Zealand Subduction Zones. Geophysical Research Letters, 23(13), 1557-1560.
Yang, T., Li, B., Fang, L., Su, Y., Zhong, Y., Yang, J., . . . Xu, Y. (2022). Relocation of the Foreshocks and Aftershocks of the 2021 Ms 6.4 Yangbi Earthquake Sequence, Yunnan, China. Journal of Earth Science, 33(4), 892-900.
Zhu, L. P., & Helmberger, D. V. (1996). Advancement in source estimation techniques using broadband regional seismograms. Bulletin of the Seismological Society of America, 86(5), 1634-1641.
Zhu, L. P., & Rivera, L. A. (2002). A note on the dynamic and static displacements from a point source in multilayered media. Geophysical Journal International, 148(3), 619-627.
Chinese references
Guo X. Y., Jiang C. S., Wang X. S., Tian X. (2017). Characteristics of Small to Moderate Focal Mechanism Solutions Stress Field of the Circum-Ordos Block. Journal of Geodesy and Geodynamics. 37(7), 675-685.
Hou L. F., Yan Y. X., Yang F. P., Yu G. Q., Li L. Y. (2018). Research on Deep Crustal Structure in Shanxi Reservoir Area. Journal of Geodesy and Geodynamics. 38(8), 818-822.
Li J. L., Zou Z. X., Zhang Z. F. (2007). The earthquake location for Shanxi Reservoir earthquake swarms. Seismological and Geomagnetic Observation and Research. 28(1), 14-18.
Ma Z. J., Zhong Y. Y, Han Y. B., et al. (2016). The Tectonic Conditions of Shanxi Reservoir Induced Earthquake in Wenzhou. Earth Science-Journal of China University of Geosciences. 41(8), 1413-1423.
Meng W., Guo X. Y., Li Y. H., Han L. B., Zhang Z. Y. (2022). Tectonic stress field and dynamic characteristics in the northeastern margin of the Tibetan Plateau. Chinese Journal of Geophysics. 65(9), 3229-3251.
Pei W. L., Yang Z. S., Zhou S. Y. (2022). Source Mechanisms of Small Earthquakes and Tectonic Stress Field in Xiaojiang, Yunnan. Acta Scientiarum Naturalium Universitatis Pekinensis.
Wan Y. G. (2019). Determination of center of several focal mechanisms of the same earthquake. Chinese Journal of Geophysics. 58(4), 602-608.
Yang F. P., Yu G. Q., Hou L. F., Zhou X. (2019). The Application of Magnetotelluric Sounding (MT) to Research of Reservoir-induced Earthquake in Shanxi Region. Chinese Journal of Engineering Geophysics. 16(2), 203-210.
Yang F. P., Zhao L. F., Yu G. Q. et al. (2021). Research on the Three-Dimensional Structural Characteristics and Deep Reservoir-Induced Seismic Environment in Shanxi Reservoir Area. Journal of Geodesy and Geodynamics. 41(12), 1264-1270.
Zhang F., Zhu X. Y., Zhong Y. Y. (2013). Study on the Characteristic of Focal Mechanism Solutions of the Shanxi Reservoir Earthquake. China Earthquake Engineering Journal. 35(3), 684-691.
Zhong Y. Y., Zhang F. Zhao D. (2011). Precise Relocation and Seismogenic Structure of the Shanxi Reservoir Earthquake Sequence in Wenzhou, Zhejiang Province. Journal of Seismological Research. 34(2), 158-165.
Zhong Y. Y. Zhang Z. F. Kan B. X. (2010). Simultaneous inversion of earthquake relocation and velocity structure in the Shanxi-reservoir, Wenzhou. Earthquake Research in China. 26(3), 265-272.
Zhong Y. Y., Zhu X. Y., Zhang Z. F., Zhang F. (2017). The Shanxi Reservoir Earthquake in Feiyun River. Zhejiang University press.
Zhu X. Y., Zhang F., Yu J. Y. (2010). Study on precise positioning and structure of the Shanxi reservoir earthquake, Zhejiang Province. Earthquake Research in China. 26(4), 380-390.

No competing interests reported.

Download PDF

Editor assigned by journal
29 Jul, 2024
Submission checks completed at journal
29 Jul, 2024
First submitted to journal
29 Jul, 2024

You are reading this latest preprint version

Characteristics and Seismogenic Background of the 2014 Earthquake Sequence around Shanxi Reservoir in Eastern China

Status:

Version 1

Abstract

Figures

1 Introduction

2 Data and methods

2.1 Data

2.2 Relocation method

2.3 Methods for Focal Mechanism

3 Analysis of calculation results

3.1 Relocation results

3.2 Focal mechanism results

4 Discussion

4.1 The seismogenic structure

4.2 Causes of earthquakes

4.3 Reliability of results

5 Conclusions

Declarations

Author Contribution

References

Additional Declarations

Status:

Version 1