Depth segmentation of the seismogenic continental crust: The 2008 and 2009 Qaidam earthquakes

Abstract

[1] The seismic hazard in the immediate vicinity of an earthquake is usually assumed to be reduced after rupture of a continental fault, with along-strike portions being brought closer to failure and aftershocks being significantly smaller. This period of reduced hazard will persist as strain re-accumulates over decades or centuries. However, this is only realised if the entire seismogenic layer ruptured in the event. Here we use satellite radar measurements to show the ruptures of two M w 6.3 earthquakes, occurring in almost the same epicentral location ten months apart in the Qaidam region, China, were nearly coplanar. The 2008 earthquake ruptured the lower half of the seismogenic layer, the 2009 event the upper half. Fault segmentation with depth allows a significant seismic hazard to remain even after a moderate and potentially devastating earthquake. This depth segmentation possibly exists in the case of the 2003 Bam earthquake where satellite radar and aftershock measurements showed that it ruptured only the upper half of the 15-20 km deep seismogenic region , and that the lower, unruptured part may remain as a continuing seismic hazard.

Deformation and Source Parameters of the 2015 Mw 6.5 Earthquake in Pishan, Western China, from Sentinel-1A and ALOS-2 Data

“…Additionally, from the static Coulomb failure stress changes calculated from both the uniform slip model and the distributed slip model, the area of shallower asperity suffers increased stress changes, which indicates that the slip in the shallow layer may have arisen from postseismic afterslip or aftershocks triggered by the deeper asperity ( Figure 7). If the accumulated strain energy in the shallower section is released by the afterslip or aftershocks, there will be no second, moderate earthquake shortly thereafter in the shallower section, such as the case in the 2008 and 2009 Qaidam Mw 6.3 earthquake sequences [56].…”

Fault mechanics and post-seismic deformation at Bam, SE Iran

“…Earthquakes that ruptured the top and bottom half of the seismogenic layer in separate events have been documented elsewhere (Berberian et al 2001;Elliott et al 2011) and can be explained by loading of the unruptured part of the fault following the earlier earthquake. Of these examples, the Qaidam earthquakes were separated by 10 months (Elliott et al 2011), and the Golbaf-Sirch earthquakes by 17 yr (Berberian et al 2001), showing that failure of the loaded half of the seismogenic layer can occur on timescales much shorter than the recurrence interval for faults in central and eastern Iran, which is generally >1000 yr (Fattahi et al 2006;Walker et al 2010). …”

The 2011 Mw 7.1 Van (Eastern Turkey) earthquake

“…The possibility of similarsized earthquakes occurring in almost the same epicentral location relatively soon after the first (i.e., before a typical seismic cycle would be complete) due to segmentation of the fault with depth is less well documented. This scenario was shown to be the case for two Mw 6.2 reverse faulting events which occurred 10 months apart in 2008 and 2009 at almost the same location in the Qaidam basin, China [Elliott et al, 2011]. Similarities between the Van earthquake and the deeper reverse faulting Qaidam event in 2008 that was followed by the up-dip shallow rupture of the same magnitude the following year raise concern for future seismic hazard near Van.…”

“…[42] Vertical separation of slip within the seismogenic crust and a delay in rupture between two along-dip portions of a fault are known to have happened recently in the case of a pair of M w 6.2 reverse faulting events in the Qaidam basin, NE Tibet [Elliott et al, 2011]. The first earthquake occurred Figure 15.…”