A more accurate relocation of the 2013 M s7.0 Lushan, Sichuan, China, earthquake sequence, and the seismogenic structure analysis

We use a combined earthquake location technique to relocate the M (s)7.0 Lushan, Sichuan, China, earthquake sequence of April 20, 2013. A stepwise approach, employing three existing location methods (the HYPOINVERSE method, the Minimum 1-D model, and the Double Difference method), is used to improve location precision by iteratively revising the velocity model station corrections, and hypocenter relocations throughout the process. Our stepwise approach has significantly improved the location precision of the Lushan earthquake sequence, yielding hypocenter locations with final errors of 359, 309, and 605 m in the E-W, N-S, and vertical directions, respectively, with average travel time residuals of 0.12 s. Furthermore, we analyzed the seismogenic structure surrounding the Lushan earthquake sequence by combining the results of the relocated hypocenter distribution with new focal mechanism solutions and information from regional geological and geophysical investigations. From our analysis, we conclude that the vast majority of the aftershocks of the Lushan earthquake sequence occurred at depths of 6-9 km, near the front of the southwestern segment of the NE-trending Longmenshan fault zone. Densely aligned hypocenters clearly suggest that the seismogenic structure of the mainshock consists of a set of basal thrust faults dipping to the NW at 40-50A degrees, at a ramp of the deep basal d,collement-thrust system at depths of 7-18 km. Focal mechanism solutions suggest that the seismogenic faults have produced almost pure thrusting. At least one SE-dipping back-thrust is also observed within the basement, as indicated by the hypocenter relocations, which points to either a secondary rupture plane during the mainshock or a plane of aftershock slips. A small number of minor events in the Lushan sequence are located at depths of 0-6 km, with a distribution suggesting that the three NE-trending faults with surface traces running through or passing close to the aftershock area are confined to the upper Mesozoic sedimentary cover, making them independent of the deeper thrust faults that ruptured during the mainshock. Therefore, the 2013 M (s)7.0 Lushan earthquake was a blind thrust fault generated on active thrust faults within the basement of the southwestern Longmenshan fault zone, with an upper limit estimation of the rupture length, average down-dip width, and rupture area of 40, 16, and 640 km(2), respectively.