Locating the Origin of Scattered Waves By Simulating Time Reversal of the Seismic Wavefield

Abstract:

The Source Physics Experiment (SPE) is a series of underground chemical explosions at the Nevada National Security Site (NNSS) that are improving our physical understanding how explosion sources generate seismic waves. Better understanding the origin of S-waves from explosions is a primary goal of the SPE. Even at distances of a few kilometers from the SPE sources, seismic recordings include arrivals of unknown origin that could originate as S-waves at the explosive source or from topographic and subsurface scatterers. Back propagation of time reversed seismograms has been used to determine the location of seismic events (e.g. Tromp et al., 2005; Larmat et al., 2006), and Myers et al. (2007) demonstrated that the time-reversal method can be used to determine the origin of direct and scattered waves in seismic simulations. In this study we identify the origin of distinct features in synthetic seismograms that are generated by elastic, finite-difference simulation of seismic propagation from SPE explosions through a model that has been developed specifically for the SPE. The SPE model includes 3-dimensional velocity discontinuities at geologic boundaries, as well as free-surface topography. Although the largest arrivals in the synthetic seismograms are expected to originate at the explosion source, other prominent features are likely to originate as scattered energy from model discontinuities. Scattering sources in the SPE model that are needed in order to match synthetic seismograms to field recordings of SPE shots will be identified. Conversely, model structures may be removed if they result in disagreement between synthetic seismograms and field recordings. Ultimately, we plan to constrain the origin of prominent features in field recordings of SPE shots by directly using the field recordings as inputs to time reversal simulations. Direct use of field recordings will require development of methods that account for the uncertainty of the seismic model through which time-reversed signals are propagated. This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.