The Democratic People’s Republic of Korea (DPRK) conducted its sixth announced nuclear test on 3 September 2017 03:30 UTC (mb 6.1). There was an aftershock 8.5 min later (mb 4.1). Such aftershocks were noteworthy and often associated with post–nuclear explosion collapses. We performed moment tensor (MT) and network sensitivity solution (NSS) analysis using regional long‐period surface‐wave and first‐motion (FM) polarities. We also extended this analysis to the previous five DPRK nuclear tests. The NSS results, which include MT solution uncertainties, show large isotropic components for the events and are within the population of other U.S. nuclear and collapse events on the fundamental lune. The FM data improved the NNS source‐type resolution. The agreement between MT seismic moments and independent coda‐envelope amplitudes indicated no biases with the Earth model error or poorly constrained seismic moment for shallow seismic sources. The MT for the collapse is not a pure implosion and consistent with an equivalent tensional closing crack mechanism and two‐sided vertical point force.

The DPRK aftershock has similar circumstances to the collapse 21 min after the 5 August 1982 Atrisco nuclear test. We calculated a range of volume reduction of 1.06×1054.23×105  m3 due to 2.68×10151.07×1016  N·m seismic moment range based on diorite rock properties and a closing tensile crack model. In comparison, the cavity radius‐yield scaling relation resulted in 2.84×1051.14×106  m3 volume range. The overlap in the volume range suggests that the aftershock can be explained by collapse of the explosion cavity. A less likely tunnel‐collapse scenario requires a 3‐ to 13‐km‐long tunnel to match an equivalent volume change (assuming a 30  m2 cross‐sectional area) and would possibly resulted in secondary sources and mechanism asymmetries detectable by MT methods.

You do not have access to this content, please speak to your institutional administrator if you feel you should have access.