Are we missing earthquakes? A new insight into the M 8 1556 Huaxian earthquake

Thursday, 18 December 2014
Yu Zhou1, Ji Ma2, Richard T Walker1, Xijie Feng2, Xiaogang Song3 and Barry Parsons1, (1)University of Oxford, Oxford, United Kingdom, (2)Earthquake Administration of Shaanxi Province, Xi'an, China, (3)Institute of Geology, China Earthquake Administration, Beijing, China
The M 8 1556 Huaxian earthquake that occurred in the Weihe graben, Shaanxi Province, China, is the deadliest earthquake in recorded history, with an estimated death toll of 830,000. Despite extensive damage reports from government documents of the Ming dynasty, the geological story of this normal-faulting earthquake is relatively poorly described. A discontinuous ~8 metre scarp along the frontal faults of the Weinan Platform and Huashan Range may have formed in the 1556 earthquake, but whether it did so, or whether it is a cumulative scarp formed in multiple events, is a question that is key to determining the recurrence interval and rates of active faulting. In this study we combine geodetic and geological observations in order to answer this question.

We have used Pleiades stereo imagery to produce a 1-metre resolution digital elevation model (DEM) of the area where the Chishui River cuts through the eastern segment of the Weinan Platform frontal fault. More than three thousand GPS measurements were collected in the field using network Real Time Kinematic GPS in order to calibrate and assess the Pleiades DEM. After making use of a few ground control points, the mean difference between the Pleiades DEM and GPS is 0.1 metres with a standard deviation of 0.9 metres. At least three river terraces (T1, T2 and T3) can be identified on the DEM. The youngest terrace (T1) postdates the last fault movement. Offset of the T2 forms the 8 metre scarp; the older terrace (T3) is offset by ~13 metres across the fault. Charcoal was collected from ancient soil and loess layers that overlie gravels in all three terraces.

The surface geomorphology, along with a fault exposure in a nearby quarry excavation, allows us to conclude that the 8-metre scarp was formed in a single earthquake. Furthermore, it appears that the 1556 event was preceded by a rupture of similar magnitude, sometime in the period 2000-6000 ka. The Huaxian fault is a relatively slow-moving fault, and in the Holocene has failed in large-magnitude, large stress-drop earthquakes separated by long intervals of time.