Bruce W. Patton

Anomaly Detection in Radiation Sensor Data With Application to Transportation Security

In this paper, we present a new approach for detecting trucks transporting illicit radioactive materials using radiation data. The approach is motivated by the high number of false alarms that typically results when using radiation portal monitors. Our approach is a three-stage anomaly detection process that consists of transforming the radiation sensor data into wavelet coefficients, representing the transformed data in binary form, and detecting anomalies among data sets using a proximity-based method. The approach is evaluated using simulated radiation data, and the results are encouraging. From a transportation security perspective, our results indicate that the concomitant use of gross count and spectroscopy radiation data improves identification of trucks transporting illicit radioactive materials. The results also suggest that the use of additional heterogeneous data with radiation data may enhance the reliability of the detection process. Further testing with real radiation data and mixture of cargo is needed to fully validate the results.

Multiagency Urban Search Experiment Detector and Algorithm Test Bed

In order to provide benchmark data sets for radiation detector and algorithm development, a particle transport test bed has been created using experimental data as model input and validation. A detailed radiation measurement campaign at the Combined Arms Collective Training Facility in Fort Indiantown Gap, PA (FTIG), USA, provides sample background radiation levels for a variety of materials present at the site (including cinder block, gravel, asphalt, and soil) using long dwell high-purity germanium (HPGe) measurements. In addition, detailed light detection and ranging data and ground-truth measurements inform model geometry. This paper describes the collected data and the application of these data to create background and injected source synthetic data for an arbitrary gamma-ray detection system using particle transport model detector response calculations and statistical sampling. In the methodology presented here, HPGe measurements inform model source terms while detector response calculations are validated via long dwell measurements using 2”

A Methodology for Determining the Concentration of Naturally Occurring Radioactive Materials in an Urban Environment

\times 4

Systematic Assessment of Neutron and Gamma Backgrounds Relevant to Operational Modeling and Detection Technology Implementation

”

Computational Assessment of Naturally Occurring Neutron and Photon Background Radiation Produced by Extraterrestrial Sources

\times 16

Hybrid Monte Carlo/Deterministic Methods for Accelerating Active Interrogation Modeling

” NaI(Tl) detectors at a variety of measurement points. A collection of responses, along with sampling methods and interpolation, can be used to create data sets to gauge radiation detector and algorithm (including detection, identification, and localization) performance under a variety of scenarios. Data collected at the FTIG site are available for query, filtering, visualization, and download at muse.lbl.gov.

Fast Spectrum Molten Salt Reactor Options

Abstract The detection of radioactive sources in an urban setting is greatly complicated by natural background radiation, which emanates from various materials including roadways, sidewalks, soil, and building exteriors. The method presented and demonstrated here represents an effort to characterize the concentration of naturally occurring radioactive material (NORM) in these types of materials. The location surveyed in this work was the Fort Indiantown Gap Combined Arms Collective Training Facility in Lebanon County, Pennsylvania. Over 70 measurements with a high-purity germanium detector were performed to ascertain the NORM concentrations present in the soil, asphalt, gravel, concrete, and walls found throughout the site. Monte Carlo radiation transport calculations were used to obtain detector responses for these various geometries and materials to convert these measurements into NORM concentration estimates. Finally, synthetic spectra were simulated using the predicted source terms and compared to actual measurements, showing acceptable agreement.

Development and Testing of the VITAMIN-B7/BUGLE-B7 Coupled Neutron-Gamma Multigroup Cross-Section Libraries

This report summarizes the findings of a two year effort to systematically assess neutron and gamma backgrounds relevant to operational modeling and detection technology implementation. The first year effort focused on reviewing the origins of background sources and their impact on measured rates in operational scenarios of interest. The second year has focused on the assessment of detector and algorithm performance as they pertain to operational requirements against the various background sources and background levels.

Investigations of active interrogation techniques to detect special nuclear material in maritime environments: Standoff interrogation of small- and medium-sized cargo ships

Abstract A computational assessment of the variation in terrestrial neutron and photon background from extraterrestrial sources is presented. The motivation of this assessment is to evaluate the practicality of developing a tool or database to estimate background in real time (or near–real time) during an experimental measurement or to even predict the background for future measurements. The extraterrestrial source focused on during this assessment is naturally occurring galactic cosmic rays (GCRs). The MCNP6 transport code was used to perform the computational assessment. However, the GCR source available in MCNP6 was not used. Rather, models developed and maintained by NASA were used to generate the GCR sources. The largest variation in both neutron and photon background spectra was found to be caused by changes in elevation on Earths surface, which can be as large as an order of magnitude. All other perturbations produced background variations on the order of a factor of 3 or less. The most interesting finding was that ~80% and 50% of terrestrial background neutrons and photons, respectively, are generated by interactions in Earths surface and other naturally occurring and man-made objects near a detector of particles from extraterrestrial sources and their progeny created in Earths atmosphere. This assessment shows that it will be difficult to estimate the terrestrial background from extraterrestrial sources without a good understanding of a detectors surroundings. Therefore, estimating or predicting background during a measurement environment like a mobile random search will be difficult.

Investigation of active interrogation techniques to detect special nuclear material in maritime environments: Boarded search of a cargo container ship

Abstract The potential for smuggling special nuclear material (SNM) into the United States is a major concern to homeland security, so federal agencies are investigating a variety of preventive measures, including detection and interdiction of SNM during transport. One approach for SNM detection, called active interrogation, uses a radiation source, such as a beam of neutrons or photons, to scan cargo containers and detect the products of induced fissions. In realistic cargo transport scenarios, the process of inducing and detecting fissions in SNM is difficult due to the presence of various and potentially thick materials between the radiation source and the SNM and the practical limitations on radiation source strength and detection capabilities. Therefore, computer simulations are being used, along with experimental measurements, in efforts to design effective active interrogation detection systems. The computer simulations primarily consist of simulating radiation transport from the source to the detector region(s). Although the Monte Carlo method is predominantly used for these simulations, difficulties persist related to calculating statistically meaningful detector responses in practical computing times, thereby limiting their usefulness for design and evaluation of practical active interrogation systems. In previous work, the benefits of hybrid methods that use the results of approximate deterministic transport calculations to accelerate high-fidelity Monte Carlo simulations have been demonstrated for source-detector-type problems. In this work, hybrid methods are applied and evaluated for three example active interrogation problems. Additionally, a new approach is presented that uses multiple goal-based importance functions depending on a particle’s relevance to the ultimate goal of the simulation. Results from the examples demonstrate that the application of hybrid methods to active interrogation simulations dramatically increases their calculational efficiency.