Location and Magnitude Estimation of the 9 October 2006 Korean Nuclear Explosion Using the Southern Great Basin Digital Seismic Network as a Large-Aperture Array

Abstract

Listen

Translate article

The Southern Great Basin Digital Seismic Network (SGBDSN) has been designed for monitoring high-frequency (1–40-Hz) local events at and near Yucca Mountain, Nevada, the designated site for the national high-level nuclear-waste repository. We find that the network is also effective as a large-aperture teleseismic array for monitoring events in and close to North Korea, the recent location of an underground nuclear test that occurred on 9 October 2006, 01:35:28 coordinated universal time (UTC), National Earthquake Information Center (NEIC) m b 4.3. We explain this by (1) low ambient noise, (2) energy-efficient propagation paths (the nuclear explosion and deep earthquakes on nearly the same ray path show dominant frequencies between 0.9 and 2.5 Hz), and (3) coherent signals across the SGBDSN. The network, when used as an array, provides a particularly good beam signal-to-noise ratio (SNR) for the nuclear explosion. Estimated beam SNR is 20 dB at frequencies between 0.9–2.5 Hz. Between January 1996 and December 2006, 58% (11 of 19) of the events with 7.1> m b>3.3 located within 300 km of the North Korean nuclear explosion by NEIC are considered large enough to be confidently picked by an SGBDSN analyst. The first-arrival of the North Korean event itself is apparent on 25 of the 29 SGBDSN unfiltered recordings. The direct P phase is confidently identified using slowness and back azimuth estimated with cross-correlation and frequency-wavenumber methods. Static time corrections for beamforming are estimated using 11 deep earthquakes within 300 km of the nuclear explosion. With the statistical regression model named two-way layout, we estimate the event first-arrival time offset and the mean station effects. The resulting relative time delays are used to calibrate the nuclear explosion single-array location to the NEIC location, using the delays derived from 11 earthquakes. With magnitude substituted for time, the same method is used to estimate magnitude corrections. The SGBDSN magnitude estimate is 4.3 using the Veith–Clawson body-wave magnitude formula.