ALBERT

Description: We investigate the excitation and propagation of far-field seismic waves from the 905 kg trinitrotoluene equivalent underground chemical explosion SPE-3 recorded during the Source Physics Experiment (SPE) at the Nevada National Security Site. The recorded far-field ground motion at short and long distances is characterized by substantial shear-wave energy, and large azimuthal variations in P - and S -wave amplitudes. The shear waves observed on the transverse component of sensors at epicentral distances 〈50 m suggests they were generated at or very near the source. The relative amplitude of the shear waves grows as the waves propagate away from the source. We analyze and model the shear-wave excitation during the explosion in the 0.01–10 Hz frequency range, at epicentral distances of up to 1 km. We used two simulation techniques. One is based on the empirical isotropic Mueller–Murphy (MM) ( Mueller and Murphy, 1971 ) nuclear explosion source model, and 3D anelastic wave propagation modeling. The second uses a physics-based approach that couples hydrodynamic modeling of the chemical explosion source with anelastic wave propagation modeling. Comparisons with recorded data show the MM source model overestimates the SPE-3 far-field ground motion by an average factor of 4. The observations show that shear waves with substantial high-frequency energy were generated at the source. However, to match the observations additional shear waves from scattering, including surface topography, and heterogeneous shallow structure contributed to the amplification of far-field shear motion. Comparisons between empirically based isotropic and physics-based anisotropic source models suggest that both wave-scattering effects and near-field nonlinear effects are needed to explain the amplitude and irregular radiation pattern of shear motion observed during the SPE-3 explosion.

Description: 〈span〉〈div〉ABSTRACT〈/div〉We report on high‐performance computing (HPC) fully deterministic simulation of ground motions for a moment magnitude (Mw) 7.0 scenario earthquake on the Hayward fault resolved to 5 Hz using the SW4 finite‐difference code. We computed motions obeying physics‐based 3D wave propagation at a regional scale with an Mw 7.0 kinematic rupture model generated following 〈a href="https://pubs.geoscienceworld.org/srl#rf34"〉Graves and Pitarka (2016)〈/a〉. Both plane‐layered (1D) and 3D Earth models were considered, with 3D subsurface material properties and topography interpolated from a model of the U.S. Geological Survey (USGS). The resulting ground‐motion intensities cover a broader frequency range than typically considered in regional‐scale simulations, including higher frequencies relevant for engineering analysis of structures. Median intensities for sites across the domain are within the reported between‐event uncertainties (τ) of ground‐motion models (GMMs) across spectral periods 0.2–10 s (frequencies 0.1–5 Hz). The within‐event standard deviation ϕ of ground‐motion intensity measurement residuals range 0.2–0.5 natural log units with values consistently larger for the 3D model. Source‐normalized ratios of intensities (3D/1D) reveal patterns of path and site effects that are correlated with known geologic structure. These results demonstrate that earthquake simulations with fully deterministic wave propagation in 3D Earth models on HPC platforms produce broadband ground motions with median and within‐event aleatory variability consistent with empirical models. Systematic intensity variations for the 3D model caused by path and site effects suggest that these epistemic effects can be estimated and removed to reduce variation in site‐specific hazard estimates. This study motivates future work to evaluate the validity of the USGS 3D model and investigate the development of path and site corrections by running more scenarios.〈/span〉

Description: 〈span〉〈div〉ABSTRACT〈/div〉We report on high‐performance computing (HPC) fully deterministic simulation of ground motions for a moment magnitude (Mw) 7.0 scenario earthquake on the Hayward fault resolved to 5 Hz using the SW4 finite‐difference code. We computed motions obeying physics‐based 3D wave propagation at a regional scale with an Mw 7.0 kinematic rupture model generated following 〈a href="https://pubs.geoscienceworld.org/srl#rf34"〉Graves and Pitarka (2016)〈/a〉. Both plane‐layered (1D) and 3D Earth models were considered, with 3D subsurface material properties and topography interpolated from a model of the U.S. Geological Survey (USGS). The resulting ground‐motion intensities cover a broader frequency range than typically considered in regional‐scale simulations, including higher frequencies relevant for engineering analysis of structures. Median intensities for sites across the domain are within the reported between‐event uncertainties (τ) of ground‐motion models (GMMs) across spectral periods 0.2–10 s (frequencies 0.1–5 Hz). The within‐event standard deviation ϕ of ground‐motion intensity measurement residuals range 0.2–0.5 natural log units with values consistently larger for the 3D model. Source‐normalized ratios of intensities (3D/1D) reveal patterns of path and site effects that are correlated with known geologic structure. These results demonstrate that earthquake simulations with fully deterministic wave propagation in 3D Earth models on HPC platforms produce broadband ground motions with median and within‐event aleatory variability consistent with empirical models. Systematic intensity variations for the 3D model caused by path and site effects suggest that these epistemic effects can be estimated and removed to reduce variation in site‐specific hazard estimates. This study motivates future work to evaluate the validity of the USGS 3D model and investigate the development of path and site corrections by running more scenarios.〈/span〉

Description: 〈span〉〈div〉ABSTRACT〈/div〉To improve the understanding of 〈span〉S〈/span〉‐wave generation from an explosion, a temporary deployment of 996 geophones, including both one‐component (Z) and three‐component sensors (3C), was installed from 15 April to 23 May 2016 at the Nevada National Security Site (NNSS). Sensor spacing varied from 25 to 100 m and consisted of 500 Z and 496 3C 5‐Hz geophones. Data were continuously recorded during the deployment at low gain (0 dB) from 15 April to 28 April and high gain (36 dB) from 29 April to 23 May. A buried (76.5 m depth) 5035 kg trinitrotoluene (TNT) equivalent chemical explosion (Source Physics Experiments [SPE]‐5) was recorded on 26 April. It was situated in a weathered granite body surrounded by volcanic tuffs, Paleozoic carbonates, and alluvium. The array was deployed ∼400–3000  m from the explosion. A set of large weight drop shots (13,000 kg source) at 53 locations both inside and outside the geophone array were also recorded, as were local, regional, and teleseismic earthquakes. Data recovery was good, with 95% of data recovered from the chemical explosion and up to 99% in the following weeks, including both the weight drop shots campaign and the continuous data. Important initial results from the deployment include estimates of the spatial correlation length of velocity heterogeneities and a higher resolution velocity model. Observations of the data and synthetics indicate that some far‐field (elastic) 〈span〉S〈/span〉‐wave energy is generated by scattering and conversion outside the near‐field (inelastic) region. Interferometric processing was conducted on a Hadoop big‐data cluster.〈/span〉

Description: The Department of Defense (DoD) uses over two million rounds of high-explosive (HE) munitions per year ( Defense Science Board Task Force, 2003 ). A small percentage does not explode, thus generating unexploded ordnance (UXO) in current range areas at a substantial rate. As these ranges are closed, the DoD becomes responsible for the environmental restoration of the affected properties. Current methods of UXO remediation are costly because of high false alarm rates. Our current research is to develop a complementary technology that will alleviate false alarm rate by detecting, classifying, and locating UXO in near real time (less than 1 minute) as a munition impacts the range. This technology will utilize an array of buried seismic sensors in a calibrated range area, along with a set of algorithms based on theoretical and applied seismology and statistical analysis. Initial field tests at three sites focused on developing concepts of the seismic and acoustic location of ordnance impacts. Our research program developed from these initial field tests has four primary objectives: 1) fully implement a wired seismic-acoustic ordnance impact location system for live fire ranges; 2) develop a system capability to discriminate high-order (HE), low-order (partially exploded), and zero-order (UXO) events; 3) reduce location error to a stringent program metric of 1–2 m; and 4) investigate the feasibility of developing a wireless implementation of the technology. This paper describes the procedures and results from follow-on tests that were conducted in two locations at the U.S. Army Aberdeen Proving Ground (APG), Maryland. These tests were used to evaluate potential seismic-acoustic methods and system configurations for a Seismic-Acoustic Impact Monitoring Assessment (SAIMA) system for mitigating UXO hazards. Significant results from this work include: 1) seismic impulses from low-order impacts were detected at distances up to 1,000 meters; 2) classification features based on measurements of the amplitude of acoustic and seismic phases produce clear discrimination between HE and UXO impacts; 3) calculated location solutions for HE and UXO impacts yield an average location error of 10–20 meters; and 4) empirical observation and waveform modeling demonstrated that surface waves dominate the signal at all distances and therefore should be the primary phase used for all components of analysis. Furthermore, these tests demonstrated the current system design, allowing further enhancements, is capable of meeting the initial research objectives (1) and (2). Future research will focus on improving system performance with refinement of the sensor-layout geometry and the detection and location algorithms through system error analyses and follow-on field testing.