Constraining Stochastic Variability of the Velocity Model Using Large-N Data

Date: 4/24/2019

Time: 06:00 PM

Room: Grand Ballroom

In an ongoing effort to refine the velocity model of the Nevada National Security Site we used the correlation between vertical component waveforms recorded by a dense 2D array of 996 stations (4.5 Hz geophones) with spacing of 25 m and 100 m, deployed at the Source Physics Experiment (SPE) Phase I site. The overall goal is to improve the simulation of seismic waves from SPE explosions on a broad frequency range (0-10Hz). Using data from the SPE-5 chemical explosion, cross-correlations of measured and synthetic data were compared to constrain the statistical properties of the stochastic velocity perturbations. The latter represent small-scale [they are not necessarily structural] heterogeneities that were added to the Yucca Flat basin model (GFM), considered as background velocity model. By trial and error, and using ground motion simulations we were able to recover sets of statistical properties of the small-scale velocity perturbations of the sedimentary basin layers. Comparisons of observed and simulated waveforms and cross-correlations decay as a function of inter-stations distance were also used to assess the enhancement of the velocity model, with and without stochastic variability.

We found that adding a depth-resolved stochastic variability of the velocity to GFM, to create the GFM+S models, improves the overall performance of ground motion simulations of SPE-5 in the desired frequency range of 0-10 Hz. The stochastic velocity fluctuations in the best performing stochastic velocity model have a horizontal correlation distance of 600m, vertical correlation distance of 120m, and a deviation of 8% from the nominal GFM velocity in the alluvium layers, and a horizontal correlation distance of 1200m, vertical correlation distance of 300m, and a deviation of 5% from the nominal GFM velocity in the underlying sedimentary layers.

This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. LLNL-ABS-765181

Presenting Author: Arben Pitarka

Authors