资源环境科技发展态势分析平台

The 3 September 2017 underground nuclear test (m(bNEIC) = 6.3) is the largest of six announced North Korean explosions. The event generated many P wave seismograms at global broadband seismic stations with good signal-to-noise ratio for periods less than similar to 5 s. Instrument deconvolution provides 435 stable broadband P wave ground displacement records in the period range 0.1 to 5.0 s. These are stacked in 26 azimuth/distance windows to average path and receiver effects. Waveform stacks and average amplitude of 4-Hz ground displacements are modeled assuming a Mueller-Murphy explosion source model for a granite source medium. Nonelastic pP delays consistent with burial depths in the mountainous source topography are considered, and explosion yield and an average constant-Q attenuation operator are estimated by fitting the waveforms. For a source depth of 750 m in heavily damaged environment, the estimated yield = 230 +/- 50 kt and t* = 0.78 +/- 0.03 s.

Plain Language Summary Seismic P wave ground motions recorded around the world for the 3 September 2017 underground nuclear test at the North Korean test site are modeled to constrain the explosion yield and the average path attenuation parameter. Broadband ground motion waveforms are modeled along with the amplitude of 4-Hz ground motions, exploring a wide range of attenuation and yield parameters along with allowing for nonelastic behavior of the P wave reflection (pP) from the free surface above the explosion, which travels down through the damage zone surrounding the explosion-produced cavity. For a source depth of 750 m, which is reasonable for location of the explosion under the mountain peak, and allowing for an extensive damage zone, the waveform modeling provides an estimated yield of 230 +/- 50 kt. Yield estimates of smaller events are made based on calibration of magnitude-yield relations using this result and range from 0.5 to 17.8 kt.