Fault strike changes can play an important role in the earthquake rupture process. Specifically, a fault bend may act as a nucleation point or a termination point for earthquake rupture. This phenomenon has been studied by numerical models that could predict surface‐rupture extent based on bend geometry but has not been well verified with field geologic observations. Therefore, to better explore rupture behavior of fault bends, more case studies of coseismic slip distribution of surface ruptures at fault bends associated with historical earthquakes are needed. Here, we reevaluate surface ruptures produced by the A.D. 1850 Xichang earthquake to determine the coseismic rupture length and slip distribution at an extensional fault bend of about 30°. The fault bend is located at the intersection between the Anninghe fault (ANHF) and Zemuhe fault (ZMHF) in the southeastern region of the Tibetan plateau. Integrated with the previous studies of surface ruptures along the ZMHF associated with the A.D. 1850 earthquake, we find new surface ruptures along the southern segment of the ANHF and suggest that the earthquake ruptured both the ANHF and ZMHF coseismically. This result indicates that an angle of 30° at an extensional bend may not effectively stop seismic ruptures. Based on the analysis of coseismic horizontal displacement of 1–1.5 and 3–5 m along the ANHF and ZMHF, we estimate slip gradient to be 15.4±3.7 and 1.7±1.6  cm/km to the right and left side of the Xichang fault bend, respectively. Though we cannot completely rule out nucleation on the ANHF, we suggest that a high slip gradient on the ZMHF with a sudden decrease at the Xichang fault bend is associated with the release of high seismic energy, which could enable coseismic rupture on the nonoptimally oriented ANHF. Moreover, based on detailed investigations, the rupture length associated with the earthquake is determined to be about 150 km, and the corresponding magnitude is suggested to be M7.6±0.3. A rupture length of 150 km for the A.D. 1850 Zemuhe earthquake is consistent with the distribution of shaking in historical records.

You do not currently have access to this article.