Regional Body-Wave Discrimination Research
Monitoring the world for potential nuclear explosions requires identifying them by their expected seismic signatures and discriminating them from earthquakes and other sources of seismic waves. Large events (approximately m{sub b} > 4.0) can often be successfully identified by the M{sub s}:m{sub b} discriminant. In order to monitor small events (approximately m{sub b}, < 4.0) short-period regional waveform data recorded within 2000 km will be needed because of poor signal-to-noise at large distances and/or long-periods. Many studies have shown that short-period (0.5-10 Hz) regional body wave phases (e.g. Pn, Pg, Sn, Lg and coda) have excellent discrimination power down to very small magnitudes when used at various nuclear tests sites. In order to broaden the application of these regional body wave techniques, we are developing size-, distance- and location-based corrections to apply to the regional data to allow wider data comparison and better discrimination performance. Building on prior work (e.g. Taylor et al. 1999, Rodgers and Walter, 2000), we are developing a revised Magnitude and Distance Amplitude Correction (MDAC) procedure. The procedure makes use of the very stable moment magnitude determinations from regional coda envelopes (see Mayeda et al, this Symposium) to provide an independent size estimate. Using a Brune (1970) style omega-squared source spectral model, we parameterize the source in terms of apparent stress and its scaling with moment. For the distance corrections we parameterize in terms of geometrical spreading, and frequency-dependent attenuation. In addition there are constants associated with velocities, densities and a phase- and frequency-dependent site effect. Using this relatively simple model we can remove much of the magnitude and distance trends from the regional data. We use a grid-search technique to explore the model space with more emphasis on removing the magnitude and distance trends than in fitting the observable spectra perfectly. Two-dimensional path effects are currently corrected by kriging the MDAC residuals but they could also be addressed through location-dependent attenuation models if available. Using earthquake and explosion data from the Nevada Test Site and India we show that the revised MDAC correction and kriging of the residuals can significantly improve phase and spectral ratio discrimination performance. In addition by correcting the regional amplitudes we allow maximum flexibility in exploring and creating new combinations of regional phases which may have even better discrimination potential. Finally the corrected regional amplitudes provide critical information on the detection of regional phases, which is another important factor in the overall estimation of discrimination capability.
- Research Organization:
- Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- W-7405-ENG-48
- OSTI ID:
- 15013192
- Report Number(s):
- UCRL-JC-138998; TRN: US200802%%1135
- Resource Relation:
- Conference: 22nd Seismic Research Symposium, New Orleans, LA, Sep 12 - Sep 15, 2000
- Country of Publication:
- United States
- Language:
- English
Similar Records
Regional Body-Wave Corrections and Surface-Wave Tomography Models to Improve Discrimination
LLNL Identification Program: Regional Body-Wave Correction Surfaces and Surface-Wave Tomography Models to Improve Discrimination