The Seismogenic Mechanism of the 2022 Ms 6.8 Luding Earthquake: Evidence from Pg -Wave Tomography

Abstract At 12:52 on 5 September 2022, an Ms 6.8 earthquake on the southeastern segment of the Xianshuihe fault zone struck Luding County, Ganze Prefecture, Sichuan Province, China. Studying the high-resolution velocity structures and their lateral variations in this area is important for exploring the tectonic background and seismogenic mechanisms of the Luding earthquake, as well as for forecasting future seismic activity. Here, we used waveform data from May 2019 to August 2022 recorded by 50 short-period seismometers installed around the Xianshuihe fault zone and manually checked the automatically picked P-wave arrival times. We ultimately obtained 77,066 travel times from 6514 earthquakes, and inversion yielded a high-resolution velocity structure model of the upper crust in our study area, with a lateral resolution of 5–8 km. The high-resolution P-wave velocity structure of the southeastern segment of the Xianshuihe fault zone was obtained using a large amount of data from dense seismic stations, and the following results were obtained. The structure of this segment features a high-velocity anomaly in the Moxi area, corresponding to the asperity associated with the Ms 6.8 Luding earthquake. The findings of the rupture propagation suggest that the velocity structure controlled the occurrence and propagation direction of the Luding earthquake rupture. The aftershock distribution consisted of three clusters. Two clusters were related to the stress concentration at the rupture ends. The other cluster was farther from the rupture zone, possibly due to high-stress accumulation near the high-velocity anomaly and the easier generation of new microfractures under the stress disturbance of the mainshock. Before the Luding earthquake, we accurately estimated the earthquake location and magnitude on the basis of the position and size of the high-velocity anomaly. This successful experience indicates the general applicability of the “rivet” model for effectively assessing earthquake hazards.