To understand shear wave generation observed during underground explosions several source-physics experiments (SPE) in fractured granitic rock mass were conducted at the Nevada National Security Site (NNSS). Fractured rock present challenges: fractures are poorly characterized and sparsely sampled; the geomechanical and geophysical properties of fractures are unknown and measured at the laboratory scale; and the spatiotemporal scale-disparity between the near source explosion physics and the far-field wave propagation physics requires considerable computing resources to simulate geophysical signatures from source-to-receiver. To build a credible model of the subsurface we integrated the geological, geomechanical and geophysical characterizations conducted at NNSS. Because detailed site characterization is limited we numerically investigated the effects of the characterization gaps on the overall response of the system. Using HPC, we performed several computational studies to identify the key geologic features that affect the most the ground motion in the near-field and in the far-field using stochastic representation of the subsurface. Using brute force Monte Carlo simulations and sampling judiciously the large hyperspace of parameters, we have probabilistically conducted several sensitivities studies on the geological, geomechanical and geophysical parameters. Such studies would help guiding site characterization efforts to provide the essential data to the modeling community.
Improved understanding of explosion generated wave motions through numerical modeling helps advancing the interpretation of seismic data for nuclear explosion monitoring (NEM) and nuclear explosion forensics. Currently, there are several significant challenges for the monitoring community, such as understanding the effects of emplacement material properties, depth of burial, damage, pre-stress and near-source heterogeneities on wave motion amplitudes, and the generation and partitioning of energy into different modes (e.g., compressional and shear waves, body and surface waves). Understanding these phenomena will improve yield estimation and event identification by accounting for predictable effects on wave motions due to source emplacement, near source heterogeneity, and path specific propagation effects. Knowledge gained on these effects can then be applied to regions where no empirical data exist, either in regions without historical explosion sources or seismic recordings or for explosions conducted under un-calibrated emplacement conditions. Numerical simulations provide a versatile tool to gain insight into the generation and propagation of wave motions, including both nonlinear and linear effects. Explosions are well known to involve the near instantaneous release of high temperature and pressure gas in a small volume of space. These high energy densities cause irreversible nonlinear behavior in the surrounding host material (rock) due to the generation and propagation of the outgoing hydrodynamic shock wave. It has been long appreciated that nonlinear response effects at the explosion emplacement have a strong impact on the observed far field seismic motions (e.g., Werth and Herbst, 1963; Perret and Bass, 1975; Murphy, 1981; Rodean, 1981; Denny and Johnson, 1991). However, a full understanding of these effects has been hampered by limitations in the knowledge of and computational requirements to represent all relevant nonlinear material response effects and to propagate waves from the source region to receivers. Advances in numerical methods and more powerful computational resources now make it possible to routinely compute the hydrodynamic response of earth materials to buried explosions with continuing improving fidelity. These studies justify optimism that explosion generated waves for other emplacement geologies and conditions can be predicted by hydrodynamic modeling and that the proposed approach can reduce uncertainties in NEM source estimates. Furthermore, to understand shear wave generation observed during underground nuclear explosions several surrogate chemical source-physics experiments (SPE) in fractured granitic rock mass were conducted at the Nevada National Security Site (NNSS). SPE data will be used to explain the genesis of the observed shear motions.