Re: “Monitoring Nuclear Weapons and Nuclear Explosive Materials”
Physics and Society, Vol. 35, No. 3. (14 July 2006)
Regarding the subject article by Steve Fetter and Ben Rusek, and thus too the associated 2005 CISAC-National Academy report, several technical means are described that would make it seem that multiple tools are in-hand to reliably and securely monitor, as part of some cooperative arms reduction agreement, nuclear weapons, dismantled components and other sensitive nuclear explosive materials. This is not quite the case, yet. While there are numerous technical approaches that would seem to have merit, implementation is very problematic, to the point of being prohibitive. It is these implementation problems that need the real attention.
There is probably no more worthy goal for physicists and other technical specialists working on nonproliferation and arms reduction problems than that of developing viable nuclear arms reduction monitoring techniques. And there is a concomitant high degree of responsibility for policymakers to support research and development of these technical means, and to educate themselves on their efficacy and security.
That said, we are not served well if we are led to believe that the technical means to solve some of the most important nuclear arms reduction monitoring and inspection issues are readily available and waiting for responsible policy makers and security specialists to use them. Examining two of the areas mentioned in the article elucidates some of the more difficult problems: use of nuclear weapons radiation signatures, and use of tags and seals.
Radiation Signatures. In a cooperative monitoring and inspection environment, radiation signatures are useful for validating declarations by the inspected party about the nature of the nuclear item being examined. This is particularly true for the fundamental task of counting warheads, for example. (Accurately verifying the numbers of warheads will become much more important in the future if the weapons states agree to further substantial reductions, and in fact may only do so if secure and effective technical means of accounting exist.) Warheads and components are often in secure containers, or their form is not particularly sight-sensitive nor unique. Radiation signatures can be useful to verify not only that an item has the basic form of a nuclear weapon, but also something about its uniqueness, and if dismantled, something identifiable about the dismantled parts.
There are two basic approaches to employing radiation signatures: template matching and fundamental attribute measurement. The article describes template matching as viable. Problems with template matching include the fact that one first has to verify the master is as declared, and the associated gamma-spectrum template contains a copious amount of weapon design information and is thus very sensitive. The TRIS system described in the article was developed for US safeguard, security, and other unilateral activities. It may not be appropriate for an arms reduction regime that requires high levels of assurance. It is probably not the right tool for cooperatively validating the reference item signature, and it is very problematic to discuss measurement system design and efficacy with a partner state when both parties are under a mandate to protect their weapons design information. Currently it is impossible, within existing legal constraints, to share weapons signatures, particularly the differences that would have to be accounted for in order to make the measurement system reliable. How is that accomplished without fear of revealing weapons design information?
The much better approach is the use of fundamental, unclassified nuclear weapons and nuclear explosive material (NEM) attributes. Some attributes can be discussed quite openly in significant detail, and the associated radiation signals generally are not as sensitive. Examples of such attributes include presence of Pu239 or U235, threshold NEM mass, Pu240/Pu239 mass ratio maximum value, lack of fissile material in oxide form, NEM configuration not consistent with powder or rubble pieces, and Americium content (age). US and Russian governmental technical specialists have been cooperatively working for more than a decade to develop, certify, and demonstrate systems based on the attribute approach.
But even using an attribute approach, a gamma radiation spectrum is recorded. If it is from a weapon, a weapon component, or (in the case of Russia) raw NEM, this spectrum is classified sensitive information. Special information protection (information barrier) techniques will most likely need to be incorporated into the design and assembly of the measurement system. But also, the measurement system will need to be owned and operated by the host country. Authentication of the measurement system results thus becomes the critical issue. How does the inspecting party verify that the complicated inspection system, owned and operated by the inspected party (because once it is used for the first time it will henceforth be a classified data acquisition system) yields valid and trustworthy results? It is this issue that must be demonstrably solved before a radiation signature system is available to monitor nuclear warheads and their associated components and materials cooperatively. The good news is that the United States and the Russian Federation are making good cooperative progress on this problem, but we are not there yet.
Tags and Seals. It is true that there are a wide variety of tags and seals that can be in principle applied to such items as launchers, warhead containers, and storage rooms, and that some can even be interrogated remotely. But there are only a very, very select few of these types of devices that can be trusted, because of their very high degree of tamper resistance, to be worth much in a nuclear arms reduction environment. Yes, the use of most any tag or seal typically brings with it the right of inspection, and therefore an inspecting party occasional on-site presence. This right should not be under-valued. But to believe that there is a wide variety of tags and seals that could be used to uniquely identify, and therefore accurately count, launchers, warheads or their containers is wrong because most are too-easily counterfeited, or too easy to remove and replace without detection. The use of any such technology in a cooperative environment will require that all the features of the tag or seal be known -- making it that much more vulnerable to tampering. There are only two passive methods that have passed muster in the US technical community and were once accepted by the US Government for use on strategic items in an arms reduction environment (investigated for the original START agreement): a tag made of a uv-cured slurry containing micaceous hematite that produced a highly unique and acceptably invulnerable light pattern that could be recorded, and a tag based on the unique intrinsic sub-surface ultrasonic reflection pattern of the interrogated item, similar to technology used in medical and NDT applications. It is the new generation of active (electrically powered) tags and seals that offer the greatest resistance to tampering using embedded cryptographic keys and other tamper sensors. However, they have a huge fundamental problem: there are no long-lived batteries or other miniature power sources for these devices. Thus, they have to be maintained too frequently (2-3 years), offering an excuse for frequent access and change-out that run counter to the whole purpose of tags or seals.
While CISAC, including the authors of the published article, and the National Academy of Science are to be truly lauded for their effort to comprehensively assess the (cooperative) monitoring of nuclear weapons and nuclear-explosive materials, the problems are not as easily solved as portrayed. In several critical areas, in contrast to what has been suggested, the problems have not yet been solved and the technologies do not yet meet the inherent standards required.
2003-present--Affiliate Professor, Henry M. Jackson School of International Studies, University of Washington, Seattle
1998-present--Participant, US-Russia Warhead Safety and Security Exchange Agreement
1986-2003--Member of the scientific staff and Sector Leader, Pacific Northwest National Laboratory, Defense Nuclear Nonproliferation Programs
1988-2003--Member/Chairman, Information Barrier Working Group, US DOE
1997-1998 --Chairman, Nuclear Warhead Radiation Signatures Peer Review Group, US DOE
1991-1993--Chairman, Tagging Laboratory Advisory Group, US DOE
1990-1992 --Executive Secretary, President’s Warhead Dismantlement and Fissile Material Control Advisory Group (Robinson Committee)