Optimization of surface wave identification and measurement

Accurate and reliable measurement of surface waves is important to Comprehe nsive Nuclear-Test-Ban Treaty (CTBT) monitoring because the M-s:m(b) discri minant and its regional variants can in many cases unambiguously identify e vents as earthquakes or explosions. Surface wave processing at the Internat ional Data Center (IDC) is designed to be completely automated and is perfo rmed using the program Maxsurf. Maxsurf searches for surface wave character istics in the expected surface wave arrival time window for all continuous long-period and broadband data in the IDC processing stream. The Prototype IDC GSETT3 Reviewed Event Bulletin (REB) now contains a very large and grow ing data set of surface wave measurements. Users of this data set need to b e aware of processing changes and calibration errors in the GSETT3 experime ntal bulletin. The prototype International Monitoring System (IMS) surface wave detection threshold is approximately one magnitude unit lower than the detection threshold of other global networks that use visual identificatio n of surface waves. Surface wave identification and measurement can be impr oved through development of regionalized earth models, phase-matched filter ing and the use of path corrected spectral magnitudes in place of M-s. Regi onalized earth models are developed through tomographic inversion of a very large data set of phase and group velocity dispersion measurements. Discri mination capability can be improved through the use of maximum likelihood m agnitudes and maximum likelihood upper bounds.