2021 Vol. 43 No. 3
2021, 43(3): 261-286.
DOI: 10.11939/jass.20210100
Abstract：
The MS6.4 （MW6.1） earthquake occurred on May 21, 2021 in Yangbi county, Yunnan, China is the main shock of a typical foreshock-mainshock-aftershock sequence. This research carried out a preliminary study focusing on tectonic background, double-difference hypocenter location, moment tensor and rupture directivity of major earthquakes, stress field and fault slip-tendency, as well as tidal effect. According to the results of moment tensor inversion and rupture directivity of the main shock and the distribution of aftershocks, it is determined that the source faults of the main shock has strike 137°, dip 75°, rake −167°. The centroid depth of the main shock is 6.0 km, and the fault ruptured unilaterally toward the south-east direction, showing pure double-couple mechanism with dominant right-lateral strike-slip and minor normal components. Relocated hypocenters show that the Yangbi earthquake sequence is located near, but significantly departs from the well-known Qiaohou-Weishan fault in the northern segment of the Honghe fault zone, demonstrating an unmapped NW-trending strike-slip fault （we named “Yangbi fault” in this paper） with some small scale conjugate faults of NE-trending. The hypocenters are distributed along the NW-trending major fault, but also exhibit clusters along the NE-trending faults. The strongest foreshocks and most of the major aftershocks were caused by rupture of NE-trending source faults. All major earthquakes show a unilateral rupture directivity. We have obtained reliable mechanism solutions of more than 20 events with MW＞3.4 in the focused area （within 15 km of the center of the Yangbi earthquake sequence）, which enabled us to be able to invert the mean stress field in the area. The principal stress shape ratio φ＝（σ2－σ3）/（σ1－σ3） is 0.46±0.17; the （azimuth, plunge） of the maximum, intermediate, and minimum principal stress axes are （188.0°±9.0°, 12.4°±7.0°）, （50°±45°, 72.1°±11.3°）, and （280.3°±7.0°, 10.4°±12.0°）, respectively. Through theoretical tidal strain and stress analysis, it is found that this seismic sequence is significantly affected by tidal modulation. The first major earthquakes of the foreshock clusters that began at 18:00−20:00 on May 18 and 19 and the main shock occurred near the peaks of tidal volumetric strain and Coulomb failure stress. Based on the focal mechanism solutions of the main earthquakes, the distributions of foreshocks and aftershocks, the facts of tidal modulation, the rupture directivity of major earthquakes, fault slip-tendency analysis, and the results of previous studies on similar seismic activities in northwestern Yunnan, we preliminarily suggest that the Yangbi earthquake sequence is significantly affected by the action of deep fluids. The first foreshock activity climax at 18:00 on 18 May, in a tensile fault step （should be permeable channel） area of the NW-trending fault, likely initiated along NE-trending faults （greater slip-tendency） by deep overpressure fluid, and then migrated to the northwest. The second climax of foreshock activity that began at night on the 19 May was concentrated near the hypocenter of the main shock. The triggering of these foreshocks and the action of deep fluid jointly promoted the activity of the NW-trending fault （smaller slip-tendency）, but the main shock was mainly promoted by the action of deep fluid.
The MS6.4 （MW6.1） earthquake occurred on May 21, 2021 in Yangbi county, Yunnan, China is the main shock of a typical foreshock-mainshock-aftershock sequence. This research carried out a preliminary study focusing on tectonic background, double-difference hypocenter location, moment tensor and rupture directivity of major earthquakes, stress field and fault slip-tendency, as well as tidal effect. According to the results of moment tensor inversion and rupture directivity of the main shock and the distribution of aftershocks, it is determined that the source faults of the main shock has strike 137°, dip 75°, rake −167°. The centroid depth of the main shock is 6.0 km, and the fault ruptured unilaterally toward the south-east direction, showing pure double-couple mechanism with dominant right-lateral strike-slip and minor normal components. Relocated hypocenters show that the Yangbi earthquake sequence is located near, but significantly departs from the well-known Qiaohou-Weishan fault in the northern segment of the Honghe fault zone, demonstrating an unmapped NW-trending strike-slip fault （we named “Yangbi fault” in this paper） with some small scale conjugate faults of NE-trending. The hypocenters are distributed along the NW-trending major fault, but also exhibit clusters along the NE-trending faults. The strongest foreshocks and most of the major aftershocks were caused by rupture of NE-trending source faults. All major earthquakes show a unilateral rupture directivity. We have obtained reliable mechanism solutions of more than 20 events with MW＞3.4 in the focused area （within 15 km of the center of the Yangbi earthquake sequence）, which enabled us to be able to invert the mean stress field in the area. The principal stress shape ratio φ＝（σ2－σ3）/（σ1－σ3） is 0.46±0.17; the （azimuth, plunge） of the maximum, intermediate, and minimum principal stress axes are （188.0°±9.0°, 12.4°±7.0°）, （50°±45°, 72.1°±11.3°）, and （280.3°±7.0°, 10.4°±12.0°）, respectively. Through theoretical tidal strain and stress analysis, it is found that this seismic sequence is significantly affected by tidal modulation. The first major earthquakes of the foreshock clusters that began at 18:00−20:00 on May 18 and 19 and the main shock occurred near the peaks of tidal volumetric strain and Coulomb failure stress. Based on the focal mechanism solutions of the main earthquakes, the distributions of foreshocks and aftershocks, the facts of tidal modulation, the rupture directivity of major earthquakes, fault slip-tendency analysis, and the results of previous studies on similar seismic activities in northwestern Yunnan, we preliminarily suggest that the Yangbi earthquake sequence is significantly affected by the action of deep fluids. The first foreshock activity climax at 18:00 on 18 May, in a tensile fault step （should be permeable channel） area of the NW-trending fault, likely initiated along NE-trending faults （greater slip-tendency） by deep overpressure fluid, and then migrated to the northwest. The second climax of foreshock activity that began at night on the 19 May was concentrated near the hypocenter of the main shock. The triggering of these foreshocks and the action of deep fluid jointly promoted the activity of the NW-trending fault （smaller slip-tendency）, but the main shock was mainly promoted by the action of deep fluid.
2021, 43(3): 287-302.
DOI: 10.11939/jass.20200051
Abstract：
Based on the seismic waveform data of 757 earthquakes recorded by 27 broadband stations of Tibet Autonomous Region seismic network from 2017 to 2019, it is for the first time that the reverse two-station method has been applied to 1 Hz Lg-wave Q value tomography study in the southern Tibetan Plateau. In this research, total 1 981 Lg-waves were intercepted from the velocity window of 3.5−2.4 km/s and the Q values of 13 543 paths were calculated. After testing the checkerboard recovery of 1°×1° and 0.5°×0.5° grids respectively, we got the tomography of the Lg-wave Q0 value of the southern Tibetan Plateau with 0.5°×0.5° resolution. The inversion result shows that there exhibits high attenuation and low Q values of Lg-wave in the southern Tibetan Plateau crust, which is highly consistent with the negative anomaly of P-wave velocity, the geothermal distribution and two rift valleys in the east of the Tibet. Therefore it is inferred that there may be widespread molten material in the crust of the southern Tibetan Plateau and two fluid-melting channels. The main channel is located between the Yadong-Gulu rift and the Sangri-Cona rift, and the secondary channel flows out along the Yarlung Zangbo suture zone. By analyzing the differences of molten material distribution on both sides of the Yadong-Gulu rift, it is considered that there are different dynamical evolution models in the front-end of collision between Indian Plate and Eurasian Plate, the dynamical evolution to the west of Yadong-Gulu rift is in accordance with the theory of shortening and thickening, and that to the east of Yadong-Gulu rift is consistent with the “pump” mode.
Based on the seismic waveform data of 757 earthquakes recorded by 27 broadband stations of Tibet Autonomous Region seismic network from 2017 to 2019, it is for the first time that the reverse two-station method has been applied to 1 Hz Lg-wave Q value tomography study in the southern Tibetan Plateau. In this research, total 1 981 Lg-waves were intercepted from the velocity window of 3.5−2.4 km/s and the Q values of 13 543 paths were calculated. After testing the checkerboard recovery of 1°×1° and 0.5°×0.5° grids respectively, we got the tomography of the Lg-wave Q0 value of the southern Tibetan Plateau with 0.5°×0.5° resolution. The inversion result shows that there exhibits high attenuation and low Q values of Lg-wave in the southern Tibetan Plateau crust, which is highly consistent with the negative anomaly of P-wave velocity, the geothermal distribution and two rift valleys in the east of the Tibet. Therefore it is inferred that there may be widespread molten material in the crust of the southern Tibetan Plateau and two fluid-melting channels. The main channel is located between the Yadong-Gulu rift and the Sangri-Cona rift, and the secondary channel flows out along the Yarlung Zangbo suture zone. By analyzing the differences of molten material distribution on both sides of the Yadong-Gulu rift, it is considered that there are different dynamical evolution models in the front-end of collision between Indian Plate and Eurasian Plate, the dynamical evolution to the west of Yadong-Gulu rift is in accordance with the theory of shortening and thickening, and that to the east of Yadong-Gulu rift is consistent with the “pump” mode.
Abstract：
In this paper, the S-wave splitting parameters of the waveform data recorded at 10 stations in the source area of the 2019 Changning MS6.0 earthquake, Sichuan, from April 25, 2013 to July 31, 2019 were measured by the particle motion discriminant method combined with the polarization analysis method. More than four effective S-wave splitting parameters are obtained at nine stations. The results show that the S-wave splitting parameters at the stations in the studied region are characterized by partition in space and variation over time. The characteristics of fast wave polarization direction in space are as follows: the predominant polarization direction of fast S-wave at three stations in the southeastern source area of the Changning earthquake is in the direction of NE, which is consistent with direction of regional principal compressive stress in the southeastern source area. In the northwestern source area, the predominant polarization directions of fast S-wave, nearly EW, is consistent with the directions of regional principal compressive stress in the northwestern source area. Due to the combining effect of crustal stress and complex fault structure, the three stations CJW, GXA and LQS all have two predominant polarization directions of fast S-wave. The polarization directions of fast S-wave change with time as follows: After the main shock, the polarization directions of fast S-wave at each station gradually tended to be convergence after the dispersion increases; the polarization directions of fast S-waves at the station CJW changed three months before the main shock, indicating that with the accumulation of the stress during the seismogenic process, the anisotropic characteristics at the station CJW are controlled mainly by stress instead of structure. As for the temporal distribution, the average normalized delay time of slow S-wave at each station decreased as the distance increasing from the main shock and aftershock dense area to stations, reflecting the stronger accumulation and releasing of stress during seismogenic process in the aftershock dense area. In addition, the normalized delay time of slow S-wave at the station CNI decreased significantly about six months before the main shock, and increased rapidly after the main shock, suggesting the accumulation of stress before the earthquake and the abrupt release of stress after the earthquake lead to the change in geometry of the micro-cracks in upper crust.
In this paper, the S-wave splitting parameters of the waveform data recorded at 10 stations in the source area of the 2019 Changning MS6.0 earthquake, Sichuan, from April 25, 2013 to July 31, 2019 were measured by the particle motion discriminant method combined with the polarization analysis method. More than four effective S-wave splitting parameters are obtained at nine stations. The results show that the S-wave splitting parameters at the stations in the studied region are characterized by partition in space and variation over time. The characteristics of fast wave polarization direction in space are as follows: the predominant polarization direction of fast S-wave at three stations in the southeastern source area of the Changning earthquake is in the direction of NE, which is consistent with direction of regional principal compressive stress in the southeastern source area. In the northwestern source area, the predominant polarization directions of fast S-wave, nearly EW, is consistent with the directions of regional principal compressive stress in the northwestern source area. Due to the combining effect of crustal stress and complex fault structure, the three stations CJW, GXA and LQS all have two predominant polarization directions of fast S-wave. The polarization directions of fast S-wave change with time as follows: After the main shock, the polarization directions of fast S-wave at each station gradually tended to be convergence after the dispersion increases; the polarization directions of fast S-waves at the station CJW changed three months before the main shock, indicating that with the accumulation of the stress during the seismogenic process, the anisotropic characteristics at the station CJW are controlled mainly by stress instead of structure. As for the temporal distribution, the average normalized delay time of slow S-wave at each station decreased as the distance increasing from the main shock and aftershock dense area to stations, reflecting the stronger accumulation and releasing of stress during seismogenic process in the aftershock dense area. In addition, the normalized delay time of slow S-wave at the station CNI decreased significantly about six months before the main shock, and increased rapidly after the main shock, suggesting the accumulation of stress before the earthquake and the abrupt release of stress after the earthquake lead to the change in geometry of the micro-cracks in upper crust.
2021, 43(3): 321-337.
DOI: 10.11939/jass.20200176
Abstract：
The MW9.0 earthquake on March 11, 2011 occurred in the northeastern sea of Japan triggered a catastrophic tsunami, which had significant influences on the seismic ambient noises with ocean-lithosphere coupling. This study uses the continuous data recorded by seismic stations along the coast and inland of North America, combined with the seafloor pressure data recorded by the DART system nearby, as well as the predicted tidal data to analyze the impact of this tsunami on the seismic ambient noises by using time-frequency analysis and polarization analysis methods. The results show that this tsunami had little impact on the high-frequency noise （1.3−1.5 Hz） and short-period double-frequency microseisms （0.18−0.4 Hz）, while it significantly enhanced the amplitudes of long-period double-frequency microseisms （0.1−0.15 Hz）, single-frequency microseisms （0.05−0.08 Hz）, as well as the Earth’s background free oscillations （0.004−0.007 Hz）. Moreover, as the frequency of the ambient noise decreases, the amplitude enhancement becomes more pronounced and lasts longer. When the tsunami reaches the shore, it has an impact on the ambient noise of coastal stations nearby for all three frequency-bands of Earth’s background free oscillations, microseisms, and high-frequency noise and becomes the main energy source of seismic ambient noises at all three frequency bands. And the position of the dominant tsunami sources changes with time. All above-mentioned suggests that the influence of tsunami on seismic ambient noise is related to the propagation characteristics of tsunami, that is, due to the influence of water depth, reflections and diffractions of seafloor and inshore topography, the energy accumulation area changes with time, and the energy does not propagate uniformly to the coast, resulting in the changes of dominant polarization direction of seismic ambient noise in different frequency bands with time. By cross-disciplines of seismology and oceanography, this study is of significance for advancing studies on the coupling mechanism of ocean-lithosphere, identifying the characteristics of tsunami waves from the perspective of seismology, as well as exploring new tsunami warning mechanisms.
The MW9.0 earthquake on March 11, 2011 occurred in the northeastern sea of Japan triggered a catastrophic tsunami, which had significant influences on the seismic ambient noises with ocean-lithosphere coupling. This study uses the continuous data recorded by seismic stations along the coast and inland of North America, combined with the seafloor pressure data recorded by the DART system nearby, as well as the predicted tidal data to analyze the impact of this tsunami on the seismic ambient noises by using time-frequency analysis and polarization analysis methods. The results show that this tsunami had little impact on the high-frequency noise （1.3−1.5 Hz） and short-period double-frequency microseisms （0.18−0.4 Hz）, while it significantly enhanced the amplitudes of long-period double-frequency microseisms （0.1−0.15 Hz）, single-frequency microseisms （0.05−0.08 Hz）, as well as the Earth’s background free oscillations （0.004−0.007 Hz）. Moreover, as the frequency of the ambient noise decreases, the amplitude enhancement becomes more pronounced and lasts longer. When the tsunami reaches the shore, it has an impact on the ambient noise of coastal stations nearby for all three frequency-bands of Earth’s background free oscillations, microseisms, and high-frequency noise and becomes the main energy source of seismic ambient noises at all three frequency bands. And the position of the dominant tsunami sources changes with time. All above-mentioned suggests that the influence of tsunami on seismic ambient noise is related to the propagation characteristics of tsunami, that is, due to the influence of water depth, reflections and diffractions of seafloor and inshore topography, the energy accumulation area changes with time, and the energy does not propagate uniformly to the coast, resulting in the changes of dominant polarization direction of seismic ambient noise in different frequency bands with time. By cross-disciplines of seismology and oceanography, this study is of significance for advancing studies on the coupling mechanism of ocean-lithosphere, identifying the characteristics of tsunami waves from the perspective of seismology, as well as exploring new tsunami warning mechanisms.
2021, 43(3): 338-349.
DOI: 10.11939/jass.20200112
Abstract：
Base on eigenvalue decomposition calculation, we discussed an automatic S phase picking algorithm suitable for local earthquake events. The algorithm needs few parameters, less computer resources and is easy to realize. By choosing seven time windows with different time length, it effectively reduce the S phase pick errors that caused by unreasonable time window selection. By taking 9 855 three-component records of Fujian seismic monitoring network, we test the applicability and accuracy of this algorithm. The results show that average S phase picking error by this method is （0.003±1.34） s, 79.6% of them are within 0.5 s, 4.1% of them are greater than 2.0 s, which indicates that the algorithm can meet the demands for daily S phase picking. All above-mentioned suggests that the quality of records is the most important factor that affects the accuracy of picking results, the S phase picking errors for high signal-noise ratio records are usually smaller than those for low signal-noise ratio records, and the accuracy of S phase picking for some low signal-noise ratio records will be improved after a band-pass filter preprocess.
Base on eigenvalue decomposition calculation, we discussed an automatic S phase picking algorithm suitable for local earthquake events. The algorithm needs few parameters, less computer resources and is easy to realize. By choosing seven time windows with different time length, it effectively reduce the S phase pick errors that caused by unreasonable time window selection. By taking 9 855 three-component records of Fujian seismic monitoring network, we test the applicability and accuracy of this algorithm. The results show that average S phase picking error by this method is （0.003±1.34） s, 79.6% of them are within 0.5 s, 4.1% of them are greater than 2.0 s, which indicates that the algorithm can meet the demands for daily S phase picking. All above-mentioned suggests that the quality of records is the most important factor that affects the accuracy of picking results, the S phase picking errors for high signal-noise ratio records are usually smaller than those for low signal-noise ratio records, and the accuracy of S phase picking for some low signal-noise ratio records will be improved after a band-pass filter preprocess.
2021, 43(3): 350-358.
DOI: 10.11939/jass.20200137
Abstract：
Geo-electric field and geo-electrical resistivity observation is one of the most important ways of earthquake monitoring. In recent years, geoelectric field observation has been subjected to more and more interference caused by subways, high-voltage direct current transmission, and electrical facilities, etc. Among all these interferences, there is a special type of known interference in the earthquake geo-electric field observation, which is so-called “interference from current” caused by the electric current in measuring geo-electrical resistivity if the geo-electric field and geo-electrical resistivity were observed at the same site. This kind of interference is characterized by short interference time, fixed interference waveform and fixed appearance time, and it will cause difficulties in identifying normal variation of geo-electric field and analyzing data. This paper deals with approaches of eliminating the interference data by using interpolation method. The fractal interpolation method and traditional Lagrange interpolation method were separately used. On the basis of introducing the principle of the two interpolation method, the processing result of the two methods on simulation data and actual data are compared, it demonstrates that fractal interpolation method has higher accuracy than that of Lagrange interpolation method. Then the fractal interpolation method was used in the actual data processing. The result shows that the method not only retrieves the section information effectively, but also preserves the overall original variation tendency of the observation data.
Geo-electric field and geo-electrical resistivity observation is one of the most important ways of earthquake monitoring. In recent years, geoelectric field observation has been subjected to more and more interference caused by subways, high-voltage direct current transmission, and electrical facilities, etc. Among all these interferences, there is a special type of known interference in the earthquake geo-electric field observation, which is so-called “interference from current” caused by the electric current in measuring geo-electrical resistivity if the geo-electric field and geo-electrical resistivity were observed at the same site. This kind of interference is characterized by short interference time, fixed interference waveform and fixed appearance time, and it will cause difficulties in identifying normal variation of geo-electric field and analyzing data. This paper deals with approaches of eliminating the interference data by using interpolation method. The fractal interpolation method and traditional Lagrange interpolation method were separately used. On the basis of introducing the principle of the two interpolation method, the processing result of the two methods on simulation data and actual data are compared, it demonstrates that fractal interpolation method has higher accuracy than that of Lagrange interpolation method. Then the fractal interpolation method was used in the actual data processing. The result shows that the method not only retrieves the section information effectively, but also preserves the overall original variation tendency of the observation data.
Abstract：
Based on the one second sampling data from fluxgate magnetometers in western Chinese mainland from 2015 to 2019, this paper carried out the vertical component polarization analysis of the frequency band between 5 s to 100 s, and then processed the analysis results by some mathematical methods. The results show that the high polarization value has no obvious shape and amplitude change in both meridional and zonal directions, and the high polarization value has nothing to do with geomagnetic field disturbance. On this basis, 18 high-value anomalies were screened out, and their spatial distribution map was obtained by interpolation method. The results show that within half a year after the synchronous appearance of multiple high-value anomalies of the vertical component polarization of geomagnetic field, the high-value region may have a strong earthquake with magnitude over M6.0. After the appearance of high value phenomenon, several high value regions are likely to have strong earthquakes; furthermore the magnitude of subsequent strong earthquake seems to be positively correlated with the area of the high-value zone.
Based on the one second sampling data from fluxgate magnetometers in western Chinese mainland from 2015 to 2019, this paper carried out the vertical component polarization analysis of the frequency band between 5 s to 100 s, and then processed the analysis results by some mathematical methods. The results show that the high polarization value has no obvious shape and amplitude change in both meridional and zonal directions, and the high polarization value has nothing to do with geomagnetic field disturbance. On this basis, 18 high-value anomalies were screened out, and their spatial distribution map was obtained by interpolation method. The results show that within half a year after the synchronous appearance of multiple high-value anomalies of the vertical component polarization of geomagnetic field, the high-value region may have a strong earthquake with magnitude over M6.0. After the appearance of high value phenomenon, several high value regions are likely to have strong earthquakes; furthermore the magnitude of subsequent strong earthquake seems to be positively correlated with the area of the high-value zone.
2021, 43(3): 376-386.
DOI: 10.11939/jass.20200138
Abstract：
When using numerical methods to synthesize ground motions, in addition to fitting the response spectrum, the fitting of peak displacements and the simulation of the non-stationary characteristics of natural ground motions also have important significance and engineering application prospects. Based on the wavelet function fitting method, this paper proposes a ground motion acceleration time history that can synthesize the target response spectrum and peak displacement at the same time. Numerical calculation examples show that this method has a faster rate of convergence, and can achieve a higher precision fitting of the target response spectrum and target peak displacement through less iterative calculations. Moreover, compared with other existing methods, the ground motion synthesized by this method can retain the non-stationary characteristics of natural ground motion, due to the local characteristics of wavelet function in time domain.
When using numerical methods to synthesize ground motions, in addition to fitting the response spectrum, the fitting of peak displacements and the simulation of the non-stationary characteristics of natural ground motions also have important significance and engineering application prospects. Based on the wavelet function fitting method, this paper proposes a ground motion acceleration time history that can synthesize the target response spectrum and peak displacement at the same time. Numerical calculation examples show that this method has a faster rate of convergence, and can achieve a higher precision fitting of the target response spectrum and target peak displacement through less iterative calculations. Moreover, compared with other existing methods, the ground motion synthesized by this method can retain the non-stationary characteristics of natural ground motion, due to the local characteristics of wavelet function in time domain.
2021, 43(3): 387-391.
DOI: 10.11939/jass.20210079
2021, 43(3): 392-396.
DOI: 10.11939/jass.20210085
Top Read
Notice
Links
Hot Search
