In an age where nuclear reactors and technologies are increasingly common, prevention of nuclear attack is a growing security concern. Previously, the U.S. regulatory approach has been to control the flow of enriched, weaponizable elements. However, as enrichment technologies have become more accessible internationally, a greater emphasis is being placed into detecting radiative material in hiding or in transit.
A critical component to detecting nuclear weaponry is catching traces of special nuclear material (SNM). Materials in this category include several isotopes of uranium and plutonium. A common characteristic of these weaponizable elements is the emission of high energy photons, also known as gamma rays. Investigators often use beams of gamma rays in combination with a radiograph, which operates similarly to a medical x-ray machine, to detect notable signatures if in the presence of SNM.
Each point on a gamma-ray spectra represents the numbers of photons emitted versus its energy level. Many isotopes have well characterized spectra that are highly distinctive. Thallium 208, for example, produces a high energy gamma ray at 2.614 MeV. [1]
Prior to 9/11, it was often seen as acceptable to spend a month measuring and analyzing an individual spectrum. [2] However, it is increasingly important to gather reliable data within short amounts of time. The goals of modern detection technologies are to maximize the probability of detecting an actual threat, while making a reasonable amount of tradeoffs in order to control the amount of false positives. Factors that affect detection performance include: testing time, target location, the surrounding environment, and the analysis algorithm.
While gamma spectroscopy is very reliable for detecting SNM in laboratory settings, it is much harder to produce satisfactory results in an uncontrolled settings where multiple sources of radiation and noise commonly exist. The noise often comes from the statistical variation in the counts within each of the gamma ray energy bins. The relative severity of the issue rises if the count is low.
These uncertainties can be further amplified if terrorists have included shielding to accompany the weaponized material. Typically, Gamma rays will travel a few meters through low-Z material such as plastic, but may be absorbed or reflected by high-Z material. [3] As a result, the rays change their energy composition and spectrum based on the properties of the materials they pass through and shielding will add complex unknowns to the result.
Recent work in signal processing has led to several algorithms that attempt to alleviate the two aforementioned problems of noise and shielding. In April 2010, Sandia National Laboratories developed the GADRAS package to analyze radioactive materials. [2] Previously, many algorithmns focused primary on the peaks of the gamma spectrum, usually its most prominent feature. GADRAS however, utilizes data from the entire spectrum. This was based on the insight that counts in a gamma spectrum are often located outside the peaks, for example, less than 3% of the gamma counts for U-238 occur in its peak at 1.001-MeV. [2] Furthermore, peaks often overlap and cause ambiguity for the source. The counts outside the peaks provide information for the thickness and composition of the shielding material. By using the entire spectrum, and accounting for areas of the spectrum where the count is absent, the newer algorithms are much more adept at identifying radioactive material in the presence shielding and background radiation.
To combat the issue of noise and scattering, the Domestic Nuclear Detection Office is conducting tests on advanced radiography scattering mechanisms. One of these systems is the Cargo Advanced Automated Radiography System. The system consists of two components. The first produces gamma rays from material such as cobalt-60. The photons are passed through a fan-like bean through the object to be detected. A piece of high-Z metal is placed before the photon detector, with a hole aligned with the direction of detection to eliminate most of the photons that have been scattered by cargo that could obstruct small features, such as a piece of SNM. [4] The result is a higher signal-to-noise ratio that increases the ability of the detector to pick out suspicious cargo.
In conclusion, although the physical principles behind gamma spectroscopy are highly developed, the algorithms used on the data gathered, as well as techniques to apply the radiography are still being constantly improved.
© Yuxiang Zhou. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
[1] M. V. Hynes, "Denying Armageddon: Preventing Terrorist Use of Nuclear Weapons," Ann. Am. Acad. Political Soc. Sci. 607, 150 (2006).
[2] J. Medalia, "Detection of Nuclear Weapons and Materials: Science, Technologies, Observations," Congressional Research Services Report for Congress R40154, 4 Jun 10.
[3] G. P. Westphal, "High Rate Gamma Spectroscopy and Related Problems," J. Radioanal. Nucl. Chem. 61, 111 (1981).
[4] J. I. Katz et al., "X-Radiography of Cargo Containers," Science and Global Security 15, 49 (2007).