Segmentation and Recent rupture history of the Xianshuihe Fault, Southwestern China

Abstract

Tibetan Plateau in western Sichuan Province, China, is one of the world's most active faults, having produced at least 8 earthquakes of M ≥ 7 since 1725 on a 350-km-long segment of the fault. In the more limited 150-km long northern segment including Luhuo and Daofu, 5 earthquakes of M ≥ 6.9 have occurred during the past 100 years alone, with well documented overlapping surface ruptures. Some of these events were remarkably similar in magnitude and location to earlier historic events, suggesting that the characteristic earthquake model may apply here. No obvious geometric segmentation characterizes the smoothly curving Luhuo-Daofu sector, although its ends are marked by major left-stepping en echelon offsets that have also been the terminating points of individual earthquake ruptures. Minor en echelon offsets and bends within this segment are associated with local vertical relief, and one 9° bend is located close to the epicenter of the 1973 Luhuo earthquake (M = 7.6). In the southern sector of the Xianshuihe fault, south of Qianning, the faults splits into several branches, one of which broke over its entire 27-km length during the 1955 Kangding earthquake (M = 7.5). The main fault near Kangding is characterized by a series of restraining and releasing bends, associated with corresponding topography, and these bends may be related to the fact that this southern segment of the fault seems to be characterized by infrequent great earthquakes (e.g., M = 7-3/4 in 1786) rather than by more frequent moderate earthquakes (M = ± 7) such as those that characterize the straighter and more continuous Luhuo-Daofu sector to the north. Continuing creep has been documented along some segments of the fault, and this, together with its high degree of activity, superb high-altitude exposures, and other unique attributes, make the Xianshuihe fault one of the most promising sites in the world for earthquake-prediction, hazard-evaluation, and segmentation studies.

Files

Additional details