The Comprehensive Nuclear-Test-Ban Treaty: Technical Issues for the United States

Pierce S. Corden


On September 24, 1996, the United States, having taken the lead in initiating the negotiation of the Comprehensive Nuclear-Test-Ban Treaty (CTBT) in the Geneva Conference on Disarmament, was the first country to sign it. The CTBT is the culmination of over four decades of proposals, initiatives, intensive research and capacity development in test-detection technologies, and the contributions of partial test bans: a treaty outlawing all nuclear explosive testing of nuclear weapons. The U.S. saw the Treaty as imposing a substantial constraint on the nuclear weapon programs of potential adversaries while not damaging its own deterrent capabilities, and as an important component of its strategy of leadership in pursuing arms control, nonproliferation and global stability.

However, in October 1999, when the Treaty was taken up by the U.S. Senate for its advice and consent to ratification, after a hurried set of hearings by the Senate Armed Services and Foreign Relations Committees and three days of floor debate, the Senate rejected the Treaty. Concerns voiced during the proceedings focused on whether U.S. nuclear weapons can be confidently maintained without nuclear explosive testing, on whether the Treaty is effectively verifiable, on the impact of cheating that escaped detection, and on the Treaty’s value for U.S. national security as an arms control and nonproliferation measure.

In response to this stinging defeat, the Clinton administration undertook an extensive review of the Treaty to consider further the concerns voiced in the Senate. Retired Chairman of the Joint Chiefs of Staff General John Shalikashvili was asked to chair the overall effort[1]. He in turn asked the National Academy of Sciences to review technical issues related to the CTBT. The Academy’s study, carried out under the auspices of the National Research Council, was released in 2002[2]

The administration of President George W. Bush did not seek Senate reconsideration of the Treaty. However, in 2009 the Obama administration stated its intention to do so, and asked the National Academy of Sciences to review and update the technical issues dealt with in the 2002 report. In response to this request, on March 30, 2012, the National Research Council (the operating arm of the Academy) released the report of the Committee established to carry out the review and update: “The Comprehensive Nuclear Test Ban Treaty: Technical Issues for the United States.”[3]

The third nuclear test carried out by the Democratic People’s Republic of Korea (DPRK) on February 13, 2013, and the overwhelmingly negative reaction to the test as a threat to U.S. national security and to regional and global security and stability, attest to the continuing relevance of the Treaty’s objective of halting nuclear weapon testing universally.

In this article, I will briefly review the history of nuclear testing and the CTBT and the assessments made in the new report in the context of achieving the security objectives of ratification and entry into force of the Treaty.


The first nuclear weapon test explosion, code-named Trinity, took place at Alamogordo, NM, on July 16, 1945. Since then, eight states have carried out over two thousand nuclear tests, with yields ranging from a fraction of a kiloton to some 50 megatons. These tests have supported the development and deployment of a wide variety of types and great numbers of nuclear weapons. The USSR stopped testing in 1990, and the United States, after conducting over 1000 tests (about fifty percent of the global total), stopped in 1992. Since 1998, when Pakistan and India tested (India tested first in 1974), only the DPRK has tested: in 2006, in 2009, and now again in 2013. At present, the DPRK appears to be the only country intent on testing further. The near-complete set of moratoriums is a consequence of welcome movement toward less reliance on nuclear weapons after the Cold War, and a general if incomplete commitment to halting the further proliferation of nuclear weapons capabilities, both qualitative and quantitative.

The first proposal to halt nuclear testing, with a view to capping advances in nuclear weapons technology, was made in 1954 by Indian Prime Minister Jawaharlal Nehru. By 1958, international interest in test cessation led to adoption of a moratorium, and a Conference of Experts, convened in Geneva, Switzerland, outlined the components of a verification system for a complete, or comprehensive, test ban treaty. These included monitoring for radioactivity, and a network of seismometers in the territories of parties to the agreement. Further investigations pointed to greater difficulty in monitoring underground nuclear tests using seismic instruments than the Experts had expected. In particular, a cheating scenario, first described in 1959, which involved testing a bomb in a large cavity to “decouple” or reduce the size of seismic signals, became a concern.

In its CTBT proposal tabled in 1962, the U.S. insisted that seismic stations would be required on the territory of the USSR, as well as on-site inspections of suspicious seismic events, in order to ensure effective verification of the underground environment[4]. The negotiations failed, but in 1963 the Limited Test Ban Treaty (LTBT), negotiated among the United States, the United Kingdom, and the USSR, was achieved; this banned all tests save those underground. LTBT monitoring and verification are by National Technical Means (NTM). The LTBT, although opened to all states, was essentially an East-West agreement; France and China have never ratified it. [As described in Comprehensive Nuclear Test Ban Treaty, Message from the President of the United States, Treaty Doc. 105-28, U.S. Government Printing Office, Washington, 1997, the term "national technical means" of verification includes such means as nationally owned and operated seismic sensors. It also includes ground stations, ships, aircraft and reconnaissance satellites that use sensors of types other than those in the International Monitoring System. In particular, in the CTBT, national technical means can be used as a basis for requesting an on-site inspection.] 

By the time the CTBT was negotiated at the Geneva Conference on Disarmament from 1993-6, nearly four decades of research and development in seismic instrumentation and other nuclear test detection technologies had not only greatly increased the capabilities of National Technical Means, they had also laid the groundwork for the new Treaty’s verification regime. For the U.S., monitoring and verification continue to be provided for principally by NTM. But monitoring and verification capabilities now include an international network of seismic, radionuclide, hydroacoustic and infrasound sensors; and provision for on-site inspection of events that remain ambiguous to remote sensing. There are now sensors installed at 280 of the 321 sites in eighty-nine states of the CTBT’s International Monitoring System, aiming at global coverage[5]. Other states can of course also deploy their own NTM as they see best to focus on areas of particular interest.

However, as noted above, effective verification was not the only issue to be addressed in seeking Senate agreement to U.S. ratification of the CTBT. The range of issues included ensuring confidently that, without nuclear tests, United States nuclear weapons would not suffer degradation, with a consequent risk to the effectiveness of the weapons and to their role underpinning U.S. deterrent strategy. Such issues are obviously not dealt with directly by the Treaty, but it is understandable that U.S. policymakers would consider that they must be addressed. In 1995, the U.S. set in place a Stockpile Stewardship Program to ensure confidence in weapon performance. The Program formalized existing activities for ensuring confidence, which had in fact been employed during the period when nuclear testing was taking place, and added new resources to the effort, including much more powerful computers enabling better simulation of weapon performance, and new experimental facilities greatly expanding weapons-related experiments not involving nuclear explosions.

The aggregate monitoring and verification capabilities, and those of the Stockpile Stewardship Program, provide the backdrop for the NRC Committee’s report.

The NRC Committee

The Committee assembled to carry out the new 2012 study was an eminent one, including scientists and national security experts with backgrounds in monitoring technologies and nuclear weapons science and technology, and nuclear weapons operation. A Subcommittee dealing specifically with seismic monitoring issues was also established[6]. The Committee was asked to focus on four areas:  maintaining safety and confidence in U.S. nuclear weapon reliability in the absence of nuclear tests; monitoring capabilities – detection, location and identification of nuclear explosions; resource commitments to sustain nuclear weapons and both U.S. and international monitoring systems; and potential technical advances from cheating or unconstrained testing by others. These areas largely overlap the scope of the 2002 study. 

Although the report is described as a review and an update of the 2002 study, in its updating it goes substantially beyond a simple chronological account of the decade since the first report:  in its analyses and conclusions it adds to, and departs significantly from, the 2002 report.

Overall Technical Assessment

The Committee reached the following judgments regarding the above four areas of focus (somewhat paraphrased):
  • With commitment and adequate resources the U.S. has the technical capability to ensure without testing that its nuclear weapons are safe, secure, and effective for the foreseeable future;
  • U.S. technical monitoring capabilities have improved significantly during the past decade, but some operational capabilities are at risk;
  • U.S. global NTM are superior to those of the international networks, but the latter are important as a complement to NTM, and a basis for international policy action when NTM should remain classified;
  • There is the possibility of cheating, but the U.S. response to new nuclear weapons would not require U.S. testing whether or not the CTBT were in force;
  • The most plausible technically motivated circumstance pointing to U.S. test resumption would be a determination that nuclear weapons of types not previously tested were required in response to “adversarial nuclear activities.” The Treaty provides for withdrawal to deal with this situation; and
  • The U.S. can guard against technical surprise in any case if it sustains the relevant technical resources, which include scientific and other expertise.

In summarizing its conclusions, the 2012 Committee reports that “concerns about maintaining the capabilities to sustain U.S. national security into the future...are not the result of intrinsic technical limitations and are not limited by a possible future under the CTBT.”  However, the Committee was sufficiently concerned about what it learned in its review of financial and policy support for these capabilities that it says that a decision to provide the requisite support is required “whether or not the United States ratifies the CTBT.”  It believes that risks to the U.S. Stockpile Stewardship Program to maintain nuclear weapons capabilities are limited, such that “the CTBT would not prevent the United States from responding effectively if military and political decisions required development of previously tested weapon types not now present in the stockpile.”  Were new types required, the U.S. has the option to withdraw from the CTBT. The U.S. can deal with threats from secret cheating on the test ban by others  “as effectively under the CTBT as it could without the CTBT.”   Were an adversary to seek to develop a new type of strategic weapon, the U.S. would observe the required testing, and if U.S. technical capabilities are sustained the U.S. “will retain effective protection against technical surprises” regardless of CTBT ratification.

These are remarkable statements. The Committee concludes, in effect, that a decision to seek Senate consent to ratification can comfortably be taken on technical grounds. Other factors will then be of greater salience, such as how ratification will impact on U.S. objectives to reduce nuclear weapon numbers, cap the qualitative weapon capabilities of others, preserve and strengthen the Non-Proliferation Treaty and broader nonproliferation regime, and best engage in dealing with the current problematic cases of the DPRK and Iran.

Confidence in Reliability; Effective Verification and the Impact of Cheating

Before turning in more detail to these additional factors for ratification, it is of interest to consider further some of the specific matters dealt with in the Committee’s report.

Maintaining Confidence in Reliability

The Committee states that, with regard to U.S. nuclear weapons, U.S. technical knowledge and capabilities for stockpile maintenance have significantly advanced since 2002. And lifetime extension programs for warheads and bombs can be carried out effectively without nuclear explosive tests. The Committee cites the accumulated experience since 2002 in assessing that maintenance can be assured with continued supporting resources of personnel, facilities, and funding.

Of particular interest is the report’s assessment, which goes beyond  that of the 2002 report, that because of progress in the Stockpile Stewardship Program, beyond the baseline concept of warhead “refurbishment,”  “re-use or replacement of nuclear components can be considered as options for improving safety and security of the warheads.”  The 2002 approach stressed maintaining design discipline in considering modifications to the nuclear explosive package. The 2012 assessment acknowledges an increased interest in weapon safety and security after the 9/11 terrorist attacks, which might lead to a consideration of changes in the nuclear explosive package. But while expanding the envelope of options, this flexibility is constrained to considering “design changes” “within the range of U.S. tested designs.”  This means essentially interpolating between designs, not extrapolating beyond them. (Re-use refers to combining previously test-proven components from different nuclear explosive packages; replacement refers to introducing a modified component.)

This assessment is far from a blanket endorsement of implementing reuse or replacement options. It points to the critical importance of ensuring that what might be termed ‘ground truth’ in fielded weapons is maintained between previously manufactured warheads in series production and new series production. The element of expert judgment to arrive at a net assessment of confidence in the reliability of the nuclear explosive package of a nuclear warhead cannot be set aside. The question is how much of a departure from previously serially produced weapons can be accepted with the same level of assurance from expert judgment. The Committee’s view has taken into account the greatly increased capabilities to carry out detailed simulations of the nuclear weapon explosion made possible by the very large increases in computational capabilities of peta-scale computers and improved models based on better understanding of physical phenomena provided by experiments in the Stockpile Stewardship Program. These capabilities were not yet proven in 2002.

The Committee has also noted the more widespread use of the methodology of Quantification of Margins and Uncertainties (QMU) in evaluating warhead performance and retaining confidence in the warhead’s reliability. However, the Committee recommends a systematic stockpile surveillance program that is “statistically based where possible.”

The Committee also recommends a number of specific steps to retain confidence, such as continuing “experiments linked with analysis,” and ensuring appropriate weapon production facilities. With continuing support for the science and technology base, including personnel, the Committee judges U.S. nuclear weapons can be appropriately maintained without testing.

Effective Verification and the Impact of Cheating

With respect to monitoring, the Committee notes the substantial advances in capabilities, including the use of regional seismic data, and more sensitive radionuclide monitoring. Monitoring is much better than was foreseen in 2002, resulting in lower explosive yield detection levels at a given detection probability.

The Committee has helpfully presented its analyses for detection with different degrees of probability. For example, station networks are evaluated at 90% or 10% probability for three-station detection. This introduces important clarity into the typical discussion of a “threshold” for explosive yield detection, with the implication that below the threshold there is zero probability of detection. The verifying party naturally looks for high detection probability at low yield, but the potential cheater must take into account that the risk of detection has not disappeared, even at low probability, and must in addition factor in the risk of being caught by multiple systems, or surveillance for which a probability cannot be estimated. Thus the cheater can never have an absolute assurance of success.

The contribution of on-site inspections to effective verification evidently did not figure prominently in the Committee’s assessment. On-site inspections will require an in-force Treaty, and the framework for the assessment is in terms of whether or not the Treaty is in force. Verification will be even more effective under an in-force Treaty.

The Committee urges both upgrading national technical means for seismic monitoring, and sustaining the international system. In particular, NTM that rely on satellite platforms should be upgraded. There should be further research and development in radionuclide monitoring.

With so many advances in monitoring technologies and capabilities since 1958, it might be imagined that there would be little disagreement with the judgment that effective verification of the CTBT can be assured. But over time assessments of what constitutes effective verification have moved further and further toward requiring capabilities at lower and lower seismic signal magnitude. This is evidently linked to assessments of cheating that consider lower and lower yield explosions as posing a greater potential threat in their contribution to an adversary’s nuclear weapons[7].

The report therefore discusses so-called “hydronuclear” tests at some length. Such tests are described as tests in which the nuclear yield is less than that of the triggering high explosive yield. (Other ways to characterize such tests have included restricting them to nuclear explosions with a nuclear yield less than about four pounds high explosive equivalent; or including sub-critical events[8].) It considers the utility of hydronuclear testing as limited for the U.S., and although concluding that such testing could provide “some benefit” to Russia, says that this would be unlikely to lead to a Russian capability “to develop new strategic capabilities outside of its nuclear-explosion test experience.”

With regard to cheating scenarios, the Committee extends the analysis begun in the 2002 report. It concludes that monitoring advances make cheating by testing under cover of conventional mine explosions now a less credible scenario. It also concludes that with the use of regional monitoring and general advances in monitoring, a cheater, even utilizing a cavity, would be constrained to testing below one kiloton “to ensure no more than a 10 percent probability of detection for IMS [the CTBT’s International Monitoring System] and open monitoring networks.” Because this conclusion does not depend on NTM, it is likely that the detection probability at one kiloton is better than 10 percent.

The Committee discusses extensively the cheating scenario of testing in a cavity. This scenario, as noted above first discussed in 1959, has had very long legs. The Committee notes that if the scenario remains of concern, U.S. should:
  • “Apply modern computational and experimental methods to understand the decoupling process in various geologies;
  • “Identify areas such as geologic salt domes advantageous for decoupling and consider the need for additional monitoring;” and
  • “Identify indicators that a county is using – or may be planning to use – decoupling as an evasion strategy.”

In evaluating advances in nuclear weapons technology that potential adversaries might seek, the Committee judged that many of these might be pursued regardless of the in-force status of the CTBT, and that the U.S. can counter them without resuming testing. Technological advances such as modern two-stage weapons by states other than Russia and China are precluded without multiple tests, which should be detectable. For Russia and China, deploying “new types of strategic nuclear weapons that fall outside the design range of their nuclear-explosion test experience” would be unlikely in the absence of “several multi-kiloton tests to build confidence in their performance,” which tests should be detectable.

Conclusions: The Larger National Security Perspective

The NRC Committee’s report makes a persuasive technical case for continuing the existing moratoriums, and, other things being equal, for ratification of the CTBT. The “other things being equal” qualifier reflects the fact that the Committee does not make an overall policy judgment for seeking Senate advice and consent to ratification. Its assessments and recommendations are presented as valid whether or not the CTBT is in force. The larger national security framework, including broad security policy factors, can accordingly be considered from this point of departure.

The eight states whose ratifications are necessary for entry into force – the U.S., China, India, Pakistan, Egypt, Iran, Israel and the DPRK – include four that are observing moratoriums (U.S., China, India, Pakistan), three not known to test (Egypt, Iran, Israel, of which Egypt and Iran are non-nuclear-weapon states party to the Non-Proliferation Treaty), and one (the DPRK) that pursues testing activity. There appears to be essentially no prospect that the U.S. will resume testing, absent a major change in the international balance among the U.S., Russia and China. At the same time, the U.S. actively seeks to preclude further testing by the DPRK, testing that together with its launching of a satellite and its missile capabilities poses a serious problem for nonproliferation and global stability. It would also be a serious blow to the global nonproliferation regime were Iran to acquire nuclear weapons, and an even worse blow were it to begin testing them. A resumption of testing by China could lead to new nuclear weapon capabilities and an upgrading of its nuclear arsenal. Further Indian and Pakistani testing would be a serious negative development for regional and global security.

None of these prospects is attractive to U.S national security in terms of nonproliferation, achieving further weapon reductions, and global stability. In every case, ratification of the CTBT would strengthen the U.S. hand in dealing with the other states whose ratifications are required for entry into force of the Treaty, and put it on the high ground in securing those ratifications. Every step that moves the Treaty closer to entry into force strengthens the stability of the moratoriums, and enables increasing pressure to be brought to bear on the remaining non-ratifying states. Every step increases the stability of the global non-proliferation regime. U.S. ratification would give political support to the objectives of nuclear weapon reductions and the cessation of fissile material production for weapons. In particular, U.S. ratification prior to the Review Conference for the Non-Proliferation Treaty scheduled for 2015 would greatly strengthen its hand at that conference.

If the administration decides to proceed actively in a dialogue with Senators to seek Senate approval of the Treaty, the NRC Committee report provides a sound technical basis for making the case that the national security benefits of ratification greatly outweigh potential risks. It is possible to overcome the doubts about the nonproliferation value of CTBT ratification voiced in 1999, and complete the persuasive case for Senate action. A positive result of the benefit-risk calculus was nearly adopted in the early 1960s. In signing the CTBT, 183 states have endorsed the benefit side. The Treaty addresses the threats of  both “horizontal” proliferation, making acquiring a primitive nuclear weapon capability a minimalist prospect, and “vertical” proliferation, cutting off technical avenues to greater capabilities. It is key to reversing them both.


[1] Findings and Recommendations Concerning the Comprehensive Nuclear Test Ban Treaty, January 2001, available at

[2] Technical Issues Related to the Comprehensive Nuclear Test Ban Treaty, Washington, D.C. National Academy Press, (, 2002.

[3] The Comprehensive Nuclear Test Ban Treaty: Technical Issues for the United States, Washington, D.C. The National Academies Press (, 2012.

[4] International Negotiations on Ending Nuclear Weapon Tests: September 1961-September 1962, U.S. Government Printing Office: United States Arms Control and Disarmament Agency, Publication 9, released October 1962.

[5] The status of stations installed and certified or under testing in the International Monitoring System is as follows:
  • 45 of 50 primary seismic stations;
  • 113 of 120 auxiliary seismic stations;
  • 45 of 60 infrasound stations;
  • 11 of 11 hydroacoustic stations; and
  • 66 of 80 radionuclide stations.

See also S. Biegalski, “International Monitoring System of the Comprehensive Test-Ban Treaty,” Physics & Society 39(1) 9-12 (2010).

[6] The Committee included Ellen Williams, then at the University of Maryland, Chair; Marvin Adams, Texas A&M; Linton Brooks, former Administrator of the National Nuclear Security Administration; Ted Bowyer, PNNL; Donald Cobb, LANL (ret.); Richard Garwin, IBM (emeritus); Raymond Jeanloz, Univ. of Cal. Berkeley; Adm. Richard Mies, USN (ret.), and Bruce Tartar, Director Emeritus of LLNL. The Subcommittee included Lynn Sykes, Columbia, Chair; Hans Hartse, LANL; Paul Richards, Columbia; Gregory van der Vink, Terrametrics; and William Walter, LLNL.

[7] “Effective” verification recognizes that a judgment of sufficiency in verification capabilities is involved with respect to risk from undiscovered cheating. Effective verification would uncover cheating that posed a significant military threat that could not be dealt with in a timely way.

[8] Thomas B. Cochran and Christopher E. Paine, The Role of Hydronuclear Tests and Other Low-Yield Nuclear Explosions and Their Status Under A Comprehensive Test Ban: Washington, D.C., Natural Resources Defense Council, Inc., April 1995 (Rev.1).


Pierce S. Corden
Visiting Scholar, Center for Science, Technology and Security Policy
American Association for the Advancement of Science

These contributions have not been peer-refereed. They represent solely the view(s) of the author(s) and not necessarily the view of APS.