A preliminary report of the Yangbi, Yunnan, MS6.4 earthquake of May 21, 2021

Seismicity in the Yangbi area is relatively active (Figure 1). Since 1970, 145 earthquakes of magnitude greater than 3.0 have occurred within 50 km, including 108 Ms3.0−3.9 events, 27 Ms4.0−4.9 events, 9 Ms5.0−5.9 events, and the latest one reported here, which, at Ms6.0−6.9, is the strongest in this 51-year record. In the area within 100 km of Yangbi, 312 earthquakes above magnitude 3 have been recorded since 1970, including 249 Ms3.0−3.9 events, 45 Ms4.0−4.9 events, 16 Ms5.0−5.9 events, and two Ms6.0−6.9 events; the other Ms6.0 earthquake occurred in Yongsheng, Yunnan, on October 27, 2001.

Utilizing the continuous waveform data of the earthquake sequence recorded by 15 nearby broadband seismic stations, and adopting the regional earthquake full waveform fitting method (Herrmann et al., 2011;Herrmann, 2013), we calculated the focal mechanism solutions of the fore-, main-, and after-shocks of magnitude greater than M s 4.0 (Table 2 and Figure 2). Because of the interference of the mainshock coda, waveforms following in at least the first half hour were disturbed; stable focal mechanism solutions could thus not be obtained for them by the waveform fitting method.
Based on the observed aftershock activity characteristics and the focal mechanism solutions, we report the following description of this sequence: (1) According to the M-t plot of the earthquake sequence ( Figure 3) and the epicenter migration D-t plot (Figure 4) a number of foreshocks occurred in the 4 days before the mainshock, including 4 M s 4.0−4.9 events and one M s 5.0−5.9 event; the largest foreshock -the M s 5.6 event -occurred 27 minutes before the mainshock. The foreshocks took place mainly to the southeast of, and 5−10 km apart from, the mainshock ( Figure 4). Aftershocks were active in the first day after the mainshock, but in the following several days only sporadic M s ≥ 3.0 events took place. The mainshock, a unilateral fracture, was at the northwest end of the aftershock zone. 4.0-4.9 5.0-5.9 6.0-6.9 7.0-7.9 Fault Stations     (2) The moment magnitude of the mainshock is M W 6.0; the strike/dip/rake of the two nodal planes of the optimum double couple model are 45°/70°/−10° and 138°/81°/−160°, respectively, and the latter coincides with the spatial trend of the aftershock distribution and the surface trend of the nearby Weixi−Qiaohou Fault; assuming this fault to be the actual fracture plane, the mainshock therefore took place on a near-vertical strike-slip fault with a small amount of normal faulting.
(3) By waveform fitting, the best centroid depth of the mainshock is 11 km, which is slightly deeper than the initial rupture. Figure 5 shows the centroid depth variation with time for this earthquake sequence. It can be seen that, except for the mainshock, the centroid depths of the foreshocks and aftershocks are mainly between 6 and 8 km; on the whole, the centroid depths are rather shallow and within a narrow range.
(4) Most focal mechanism solutions of this earthquake sequence are generally similar to each other, being mainly of the strike-slip type, similar to that of the mainshock (Table 2). But the focal mechanisms of Event 3 (a foreshock) and Event 6 (the first aftershock) ( Table 2) are, respectively, of normal and reverse fault type, which are significantly different from the focal mechanism of the mainshock. In order to verify the reliability of the data, we adopted the jackknife method (Efron and Stein, 1981) and carried out 1000 inversions by random station selection. The result verified the stability of these two focal mechanism solutions. Most earthquakes in Table 2 have a small component of normal faulting (negative rake), indicating that the fault contains a certain portion of normal faulting component. Event 3 is at the southeast end of the foreshock sequence, corresponding to a near NS-trending normal fault. Event 6 is an M s 5.2 aftershock (M W 5.1), located at the southeast end of the aftershock zone; we conjecture that Event 6 occurred on a near EW-trending secondary reverse fault, which could explain the termination of the aftershock distribution in the southeast direction.
(5) The focal mechanisms of the earthquake sequence (except for Event 3 in Table 2) indicate that the maximum principal stress axis is nearly in the NS direction, slightly biased towards the west, which agrees with the regional stress field and ground surface deformation observations (Zheng G et al., 2017;Xu Y et al., 2020). These findings indicate that the seismogenic fault is controlled by the regional stress field.
In summary, this earthquake sequence is of the fore-main-aftershock type. Aftershock activities were rather strong, but occurred mainly within one day of the mainshock, after which occurred only sporadic M s ≥ 3.0 earthquakes. Beginning at 8 o'clock, May 22, the earthquakes migrated towards the southeast, then returned to the vicinity of the May 21 mainshock. The mainshock occurred on a steep strike-slip fault, which contains a small normal fault component; the Weixi−Qiaohou Fault near the earthquake sequence is probably the seismogenic fault. Most focal mechanism solutions of the sequence are consistent with that of the mainshock. The exception is an aftershock in the southeast section, which is of the reverse type; it occurred on a fault that may terminate the aftershocks in the SE direction. The P axes of most focal mechanism solutions of the sequence are approximately in the NS direction, which tallies with the regional stress field and ground deformation observations, indicating that the seismogenic fault is under the control of the regional stress field. The centroid depths from the mechanism solutions are distributed in a narrow range (6−11 km), indicating that the aftershocks took place mainly in a rather shallow part of the fault.  The ordinate D represents the distance of earthquake projection onto the trend line, which is calculated as follows. First find a trend line crossing the earthquake sequence distribution, which is generally the major axis (N138E) of the aftershock zone, so the distance sum of all events to this line is the minimum. Draw perpendicular lines from the earthquakes to the trend line, D is the distance between the foot of an event and the foot of the mainshock; set the D of the mainshock as 0. In the plot, 1 km is added to all Ds of the sequence so the mainshock symbol is not cut by the abscissa.