Explosion Monitoring Research Articles

Explosion Detection Using Smartphones: Ensemble Learning with the Smartphone High-Explosive Audio Recordings Dataset and the ESC-50 Dataset.

Explosion monitoring is performed by infrasound and seismoacoustic sensor networks that are distributed globally, regionally, and locally. However, these networks are unevenly and sparsely distributed, especially at the local scale, as maintaining and deploying networks is costly. With increasing interest in smaller-yield explosions, the need for more dense networks has increased. To address this issue, we propose using smartphone sensors for explosion detection as they are cost-effective and easy to deploy. Although there are studies using smartphone sensors for explosion detection, the field is still in its infancy and new technologies need to be developed. We applied a machine learning model for explosion detection using smartphone microphones. The data used were from the Smartphone High-explosive Audio Recordings Dataset (SHAReD), a collection of 326 waveforms from 70 high-explosive (HE) events recorded on smartphones, and the ESC-50 dataset, a benchmarking dataset commonly used for environmental sound classification. Two machine learning models were trained and combined into an ensemble model for explosion detection. The resulting ensemble model classified audio signals as either "explosion", "ambient", or "other" with true positive rates (recall) greater than 96% for all three categories.

Probabilistic Programming for Transportable Source Characterization and Uncertainty Quantification of the North Korean Nuclear Tests 2006–2017

Abstract We introduce a transportable technique to determine the yield and depth of burial (DOB) from seismic source spectra of underground nuclear explosions. We demonstrate this technique on the six declared North Korean nuclear tests. This approach derives source spectra in absolute units from regional phase (Pg) amplitudes by correcting the observations for geometric spreading, attenuation, and site amplification. We couple the source spectra and explosion source models with a probabilistic programming framework that integrates deep learning techniques and Bayesian modeling. This approach permits the exchange of information across various data categories to quantify both the data and model uncertainty. This technique stands out as an innovative use of broad-area propagation models, making it transportable across various geologic settings. This method proves to be effective in scenarios with diverse and/or limited observational data, even when the source depth is unknown. We present new independent estimates of absolute yield and DOB that are consistent with the prior assessments, underscoring the potential of this method in enhancing transportable nuclear explosion monitoring capabilities.

On Monitoring of mining explosions conducted on the territory of the former Semipalatinsk Test Site

Several mineral deposits located on the territory of the former Semipalatinsk Test Site (STS) conduct active blasting activity. The largest number of explosions is carried out at the Karazhyra coal mine. These explosions are recorded by stations of the IGR NNC RK network, three of them are part of the International Monitoring System. These are seismic arrays Borovoye (AS057), Makanchi (PS23), Kurchatov-Cross (AS058). The distances to the mines from these stations are 668 km, 452 km, 68 km, respectively. In total, about 2800 explosions were recorded during 19 years of observations. The analysis of the results of routine processing on operational seismic bulletins of the IGR NNC RK Data Center has shown that the field of received epicenters of explosions exceeds the size of the quarry. A detailed processing of explosions waveforms was carried out to find out the reasons of this scatter. Geotool and DTK – GPMCC software included in the NDC-in-a Box package was used to process and analyze the waveforms. It was found that back azimuth values for different regional phases Pn, Pg, Sn, Lg differ from each other and have different dispersion. There are systematic deviations of back azimuths for some phases from the true one. A clear dependence of the epicenter determination accuracy on the energy (power) of the explosion was noted. Subjective factors in the work of analysts that affect the accuracy of estimates are also found. Processing recommendations to improve accuracy were given for analysts of Kazakhstan National Data Centre (KNDC).

Quantifying the Potential of Argon Detection Capabilities for Nuclear Explosion Monitoring

Quantifying the Potential of Argon Detection Capabilities for Nuclear Explosion Monitoring

Correction to: In the nuclear explosion monitoring context, what is an anomaly?

Correction to: In the nuclear explosion monitoring context, what is an anomaly?

Seismic Characterization of Subsurface Structures at Rock Valley, Nevada

ABSTRACT The Source Physics Experiment (SPE) aims at improving explosion monitoring techniques by investigating source characteristics of chemical explosions in geologic formations. One of the critical tasks in Rock Valley Direct Comparison (RV/DC), SPE phase III, is to prepare for the main experiment by characterizing the subsurface structures at the test site. Based on the seismic data acquired during an accelerated-weight-drop (AWD) seismic survey at Rock Valley, we first pick the P-wave first-arrival travel times and derive a P-wave velocity model using the adjoint-state first-arrival travel-time tomography. We then apply reverse-time migration to the processed seismic data and obtain high-resolution images of the subsurface structures along the two main survey lines. Our migration results show several reflectors corresponding to major geologic formation boundaries. We employ a multitask machine learning model to enhance the reverse-time migration images and identify faults from these images. We find that our automatically picked faults correlate well with the locations of known faults in the region in addition to many geologically undetected faults. Our subsurface characterization results refine our understanding of the geology in this region and provide valuable velocity and structural information for RV/DC geologic model building and fault identification.

Reexamination of a 40Ca(n,α)37Ar cross section measurement from existing literature

Background 37Ar, produced from the neutron activation of 40Ca in sub-surface rock and soil, is important for nuclear explosion monitoring. Most of the cross section measurements done for the 40Ca(n,α) 37 Ar reaction are in the fast neutron region. A thermal cross section measurement was attempted in 1989, with the assumption that the 4-8 MeV neutrons would be completely thermalized with 8-cm of water. Methods The experimental setup was modeled using Monte Carlo N-Particle code to see if the assumptions for the neutron distribution were accurate. A model of a 3.0 MeV deuteron source was used to recreate the neutron source from the experiment. Additional modeling was done using SCALE to compare the predictions of 37Ar yields when using the measured cross section from 1989 compared to current nuclear data models. Results Modeling showed the neutron distribution was closer to 10% thermal and 90% fast as opposed to entirely thermalized. The additional modeling showed how the overestimation of the thermal cross section would affect the predicted yield of 37Ar. Conclusions Based on the modeling results, the assumption that all the 4-8 MeV neutrons are thermalized with 8-cm of water is not accurate. The high fraction of fast neutrons interacting with the Ca will result in an overestimation of the thermal cross section.

In the nuclear explosion monitoring context, what is an anomaly?

In the early years of nuclear explosion monitoring, experts used downwind detections with meaningful ratios of radioactive species to identify an explosion. Today’s reality is sparse networks of radionuclide monitoring stations looking for weak signals. Analysts need to discriminate between industrial background radioactivity and nuclear explosion signals, even using the detection of one isotope. Aerosol and xenon measurements potentially related to nuclear tests in 2006 and 2013 announced by the Democratic People’s Republic of Korea and from worldwide civilian background radioactivity are considered when defining radionuclide detection anomalies to objectively guide the use of limited analyst resources and reduce the possibility of not detecting nuclear explosions.

Adjoint Waveform Tomography for Crustal and Upper Mantle Structure of the Middle East and Southwest Asia for Improved Waveform Simulations Using Openly Available Broadband Data

ABSTRACT We present a new model of radially anisotropic seismic wavespeeds for the crust and upper mantle of a broad region of the Middle East and Southwest Asia (MESWA) derived from adjoint waveform tomography. The new model enables fully 3D simulations of complete three-component waveforms and provides improved fits that were not possible with previous models. We inverted over 32,000 waveforms from 192 earthquakes recorded by over 1000 openly available broadband seismic stations from permanent and temporary networks in the region with highly uneven coverage. Inversion iterations proceeded from the period band 50–100 s in six stages and 54 total iterations reducing the minimum period to 30 s. Our final model, MESWA, improves waveform fits compared to the starting and other models for both the data used in the inversion and an independent validation set of 66 events. Restitution tests indicate that the model resolves features in the central part of the model to depths of about 150 km. The new model reveals tectonic features imaged by other studies and methods but in a new holistic model of anisotropic shear and compressional wavespeeds (VS and VP, respectively) covering a larger domain with smaller scale length and amplified features. Examples include low crustal VS in the Tethyan belt and low mantle VS following divergent (Gulf of Aden, Red Sea) and transform (Dead Sea fault) margins of the Arabian plate. Low VS is imaged below Cenozoic volcanic centers of the Mecca–Madina–Nafud Line, Arabian Peninsula, and the Türkiye–Iran border region. Elevated VS tracks Makran subduction under southeast Iran with near vertical dip. MESWA could be used as a starting model for further improvements, say, using waveforms from in-country seismic networks that are not currently openly available and/or smaller-scale studies targeting a shorter period. The model could be used to improve earthquake hazard studies and nuclear explosion monitoring.

Seismic moment tensor classification using elliptical distribution functions on the hypersphere

SUMMARY Discrimination of underground explosions from naturally occurring earthquakes and other anthropogenic sources is one of the fundamental challenges of nuclear explosion monitoring. In an operational setting, the number of events that can be thoroughly investigated by analysts is limited by available resources. The capability to rapidly screen out events that can be robustly identified as not being explosions is, therefore, of great potential benefit. Nevertheless, possible mis-classification of explosions as earthquakes currently limits the use of screening methods for verification of test-ban treaties. Moment tensors provide a physics-based classification tool for the characterization of different seismic sources and have enabled the advent of new techniques for discriminating between earthquakes and explosions. Following normalization and projection of their six-degree vectors onto the hypersphere, existing screening approaches use spherically symmetric metrics to determine whether any new moment tensor may have been an explosion. Here, we show that populations of moment tensors for both earthquakes and explosions are anisotropically distributed on the hypersphere. Distributions possessing elliptical symmetry, such as the scaled von Mises–Fisher distribution, therefore provide a better description of these populations than the existing spherically symmetric models. We describe a method that uses these elliptical distributions in combination with a Bayesian classifier to achieve successful classification rates of 99 per cent for explosions and 98 per cent for earthquakes using existing catalogues of events from the western United States. The 1983 May 5 Crowdie underground nuclear test and 2018 July 20 DAG-1 deep-borehole chemical explosion are the only two explosions out of 140 that are incorrectly classified. Application of the method to the 2006–2017 nuclear tests in the Democratic People’s Republic of Korea yields 100 per cent identification rates and we provide a simple routine MTid for general usage. The approach provides a means to rapidly assess the likelihood of an event being an explosion and can be built into monitoring workflows that rely on simultaneously assessing multiple different discrimination metrics.

Unsupervised Machine Learning Clustering of Seismic and Infrasound Data Quality Metrics

Abstract Developing techniques for improving quality control (QC) schemes to catch seismic and infrasound data defects continues to be an area of active research. Selecting universal thresholds for the automation of data quality (DQ) checks is an efficient way to find QC issues, but these thresholds may not apply well to multiple stations with varying DQ characteristics. In addition, these thresholds may not catch subtle changes in DQ parameters that still indicate problems. Machine learning can be an alternative way of diagnosing QC issues. K-means clustering, an unsupervised machine learning clustering algorithm, has been effectively used in the past for geophysical pattern exploration. This study furthers k-means applications to DQ analysis through clustering on DQ metrics derived from day-long segments of nuclear explosion monitoring data. Our k-means implementation on broadband seismometer DQ metrics separately clustered mass recenters, calibrations lasting at least one hour, and days without either. Applying this technique to infrasound DQ metrics revealed clusters related to physical issues at the stations, such as missing back volume screws and the flooding of ported pipe inlets. These are both examples of QC issues that are difficult to diagnose or detect through the thresholding of metrics or by inspecting waveforms and spectra. Our results show that k-means clustering can be a useful QC tool in exploring DQ patterns to assist analyst review of station operation and maintenance. The learned knowledge from this exploration can then inform a thresholding workflow on how to tailor to individual stations, or the k-means model could classify data directly.

Reexamination of a 40Ca(n,α)37Ar cross section measurement from existing literature

Background 37Ar, produced from the neutron activation of 40Ca in sub-surface rock and soil, is important for nuclear explosion monitoring. Most of the cross section measurements done for the 40Ca(n,α) 37 Ar reaction are in the fast neutron region. A thermal cross section measurement was attempted in 1989, with the assumption that the 4-8 MeV neutrons would be completely thermalized with 8-cm of water. Methods The experimental setup was modeled using Monte Carlo N-Particle code to see if the assumptions for the neutron distribution were accurate. A model of a 3.0 MeV deuteron source was used to recreate the neutron source from the experiment. Additional modeling was done using SCALE to compare the predictions of 37Ar yields when using the measured cross section from 1989 compared to current nuclear data models. Results Modeling showed the neutron distribution was closer to 10% thermal and 90% fast as opposed to entirely thermalized. The additional modeling showed how the overestimation of the thermal cross section would affect the predicted yield of 37Ar. Conclusions Based on the modeling results, the assumption that all the 4-8 MeV neutrons are thermalized with 8-cm of water is not accurate. The high fraction of fast neutrons interacting with the Ca will result in an overestimation of the thermal cross section.

Nuclear explosion monitoring network design considerations

Design of an efficient monitoring network requires information on the type and size of releases to be detected, the accuracy and reliability of the measuring equipment, and the desired network performance. This work provides a scientific basis for optimizing or minimizing networks of 133Xe samplers to achieve a desired performance level for different levels of release. The approach of this work varies the density of sampling locations to find optimal location subsets, and to explore the properties of variations of those subsets – how crucial is a specific subset; are substitutions problematic? The choice of possible station locations is arbitrary but constrained to some extent by the location of islands, land masses, difficult topography (mountains, etc.) and the places where infrastructure exists to run and support a sampler. Performance is evaluated using hypothetical releases and atmospheric transport models that cover an entire year.Three network performance metrics are calculated: the probability of detecting the releases, the expected number of stations to detect the releases, and the expected number of samples that detect the releases. The quantitative measures support picking optimal or near-optimal network of a specific station density. If a detection probability of 90% (high) was desired for a design basis release of 1014 Bq (1% of 133Xe production from a 1 kt explosion), then a very high density would be required using today's sampling and measurement technology. If the design basis release were raised to 1015 Bq, then the station density could be lowered by a factor of 3. To achieve a location goal of three station detections on average, posited here for the first time, would also require very high station density for a release of 1014 Bq.

Electrostatic precipitator collection efficiency studies using atmospheric radon progeny as aerosol analogs for nuclear explosion radionuclides

Electrostatic precipitation (ESP) is an attractive low-powered collection mechanism for mobile and fixed aerosol detection of radionuclides (RNs) for Nuclear Explosion Monitoring (NEM). Aerosol samplers deployed in the International Monitoring System use a blower to draw air through a filter media to collect particulates. ESP-based samplers collect aerosols without a filter, which can greatly increase volumetric flow capacity per watt of power consumed. ESP-based collectors may be optimized to perform low-power mobile RN collection or to improve the air throughput of existing monitoring stations. This effort describes the use of unknown concentrations of atmospheric RNs to determine the collection efficiency of a compact ESP design. For this analysis, naturally occurring radon progeny are simultaneously collected by a single stage wire-plate ESP and a filter-based sampler with a known collection efficiency. The activity of resulting samples is measured with gamma-spectroscopy and decay corrected for analysis time offsets. RN collection efficiencies are then derived for an experimental survey of ESP operational parameters that influence the ionization, transit, and collection of aerosols. At volumetric flow rates of 1.5–2 CMM, the optimized collection efficiency was calculated as 21±2%, and slower rates around 0.5 CMM resulted in 55 ±5% collection efficiency. The monitoring performance of the ESP-based collector was assessed for a simplified nuclear explosion source term by calculating the minimal detectable concentrations of short-lived fission & activation products. Results of the study suggest that a low-power ESP is feasible for NEM at distances of 100s of km.

Investigation on atmospheric radioactivity sample association using consistency with isotopic ratio decay over time at IMS radionuclide stations

For the enhancement of the International Data Centre's products, specifically the Standard Screened Radionuclide Event Bulletin, an important step is to establish methods to associate the detections of the Comprehensive Nuclear-Test-Ban Treaty-relevant nuclides in different atmospheric radioactivity samples with the same radionuclide release to characterize its source for the purpose of nuclear explosion monitoring. Episodes of anomalously high activity concentrations in samples at the International Monitoring System radionuclide stations are used as the primary assumption for being related to the same release. For multiple isotope observations, the consistency of their isotopic ratios in subsequent samples with radioactive decay is another plausible hint for one unique release. The radioxenon observations that are associated with the nuclear test announced by the Democratic People's Republic of Korea in 2013 serve as case study to demonstrate the effectiveness of this basic approach and how the additionally associated samples improve the source location. We use two distinct puff releases, both of short duration, for the atmospheric transport modelling simulations to gain further evidence and confidence in our sample association study by identifying the air masses that link the releases to multiple samples. This basic approach will support the definition of analysis procedures and criteria for automatic sample association to be implemented in the Standard Screened Radionuclide Event Bulletin, which is of relevance for an expert technical analysis.

Characterization of a [formula omitted]-[formula omitted] coincidence system comprising a plastic scintillator and HPGe detector for radioxenon measurements under the CTBT

Measurement of the four relevant radioxenon isotopes, namely 131mXe, 133mXe, 133Xe, and 135Xe, play a key role in underground nuclear explosion monitoring for ensuring compliance with the Comprehensive Nuclear-Test-Ban Treaty (CTBT). A β-γ coincidence system that employs a plastic scintillator detector and a HPGe detector was developed in Beijing Radionuclide Laboratory. The Minimum Detectable Activity (MDA) in a 24-h measurement of the system for 131mXe, 133mXe, 133Xe and 135Xe can achieve 1.8 mBq, 1.4 mBq, 3.4 mBq and 4.1 mBq, respectively.

NTO explosive monitoring by smartphone based on chemiluminescence technique

NTO explosive monitoring by smartphone based on chemiluminescence technique

Innovation in Technology and Scientific Methods for Nuclear Explosion Monitoring and Verification: Introduction

Innovation in Technology and Scientific Methods for Nuclear Explosion Monitoring and Verification: Introduction

Finite-Frequency Kernels for Pg Wavetrains

ABSTRACT Pg waves, which propagate at high frequencies through the crust, are important for tomography and explosion monitoring, and can be among the most prominent P waves observed from small seismic events. Much of what we know about Pg propagation, however, comes from ray approximations, 1D Earth models, and other simplified treatments. For an improved understanding, we use full wave-equation modeling and the adjoint-state method to calculate 0.5 Hz Pg sensitivity kernels corresponding to a range of Earth models, source depths, and epicentral distances. The resulting Pg sensitivity kernels are in many cases dominated by diffraction effects, waveguide mode effects, and other nonray behavior. For a source at the surface observed at less than 4° epicentral distance, first-arriving Pg energy shows little sensitivity to the lower crust. Deeper events channel Pg sensitivity into whatever crustal layer the event originates. In all cases, the simple picture of Pg as a wave reverberating throughout the whole crust with similar sensitivity across upper, middle, and lower crustal layers is found to be inadequate. Other details of the sensitivity kernels can be understood in terms of reflections and conversions at the free surface, which have a greater overall effect on the Pg wavetrain than reflections or conversions at the Moho. Because P–P reflection coefficients at the surface are affected by changes in shear velocity, Pg travel times depend not only on compressional wave velocity, but also on shear-wave velocity. By comparing sensitivity kernels from 1D and 3D velocity models, we show that this strong dependence on near surface velocities, in turn, imparts strong importance to shallow 3D velocity variations. These results represent, to our knowledge, the first application of the adjoint-state method to regional Pg wavetrains.

Sound Recognition Method of Coal Mine Gas and Coal Dust Explosion Based on GoogLeNet.

To solve the problems of backward means of coal mine gas and coal dust explosion monitoring, late reporting, and low leakage rate, a sound recognition method of coal mine gas and coal dust explosion based on GoogLeNet was proposed. After installing mining pickups in key monitoring areas of coal mines to collect the sounds of the working equipment and the environment, the collected sound was analyzed by continuous wavelet to obtain its scale coefficient map. This was then imported into GoogLeNet to obtain the recognition model of coal mine gas and coal dust explosions. The test sound was obtained by continuous wavelet analysis to obtain the scale coefficient map, brought into the completed training recognition model to obtain the sound signal class, and verified by experiment. Firstly, the scale coefficient map extracted from the sound signal by continuous wavelet analysis showed that the similarity between the subjective and objective indicators of the wavelet coefficient maps of the gas explosion sound and coal dust explosion sound was higher, but the difference between these and the rest of the coal mine sounds was clearer, helping to effectively distinguish gas and coal dust explosion sounds from other sounds. Secondly, the experimental results of GoogLeNet parameters can be obtained. When the dropout parameter is 0.5 and the initial learning rate is 0.001, the recognition effect of the model established by GoogLeNet was optimal. According to the selected parameters, the training loss, testing loss, training recognition rate, and testing recognition rate of the model are all in line with expectations. Finally, the experimental recognition results show that the recognition rate of the proposed method is 97.38%, the recall rate is 86.1%, and the accuracy rate is 100% for the case of a 9:1 ratio of test data to training data, and the overall recognition effect of the proposed GoogLeNet is significantly better than that of vgg and Alexnet, which can effectively solve the problem of under-sampling of coal mine gas and coal dust explosion sounds and can meet the need for the intelligent recognition of coal mine gas and dust explosions.