“The verifi cation and compliance regime for a -free world will need to be more eff ective than any disarmament arrangement hitherto envisaged. One hundred per cent verifi cation of compliance with any international arms agreement is highly improbable. In the case of , however, the security stakes will be so high that states will not agree to disarm and to disavow future acquisition of nuclear weapons unless verifi cation reduces to a minimum the risk of non-compliance”

VERTIC BRIEF • 1 • MAY 2003 BRIEF

Verifi cation of a nuclear weapon-free world Dr Trevor Findlay Dr Trevor Findlay is Introduction Moreover, the same conjunction of good Executive Director of relationships between major states that will  e verifi cation and compliance regime for VERTIC. He was formerly permit the negotiation of a nuclear disarma- an Australian diplomat a nuclear weapon-free world will need to ment treaty will overcome many of the and Project Leader on be more eff ective than any disarmament obstacles, which today seem insurmountable, Peacekeeping and Regional arrangement hitherto envisaged. One to the construction of an appropriate veri- Security at the Stockholm hundred per cent verifi cation of compliance International Peace Research fi cation and compliance system. Institute (SIPRI) in Sweden. with any international arms agreement is highly improbable. In the case of nuclear disarmament, however, the security stakes will be so high that states will not agree to disarm and to disavow future acquisition of nuclear weapons unless verifi cation reduces Verifi cation tasks for a nuclear to a minimum the risk of non-compliance. weapon-free world Similarly, the compliance mechanism must  is paper starts with the assumption that, be as compelling as possible, providing a by the time the transition to complete nuclear high degree of assurance that any violation disarmament is imminent, Russia and the will be dealt with fi rmly and eff ectively. Both  will have reduced their arsenals to below the verifi cation and compliance systems , each through a continuing must be able to cope with the most feared strategic nuclear disarmament process, non- threat to complete nuclear disarmament: strategic reductions and limitations and/or ‘breakout’.  is is when a state party is unilateral measures.  eir remaining weapons suddenly revealed to have a previously hidden are all likely to be considered strategic: tactical nuclear arsenal or to have produced new weapons, those for short-range or battlefi eld weapons. use, will have to have been prohibited and Meeting these requirements is a tall order, the ban subject to verifi cation (itself a highly but not an inconceivable one. For a start, challenging undertaking that will set prece- a verifi cation and compliance regime for dents for verifi cation of a nuclear weapon-free total nuclear disarmament will not be world).1 constructed from scratch. It will build on Depending on the size of the remain ing practical experience of the disarmament Russian and  arsenals, the lesser nuclear process as it moves towards zero and will weapon states (China, France, India, Israel, draw on yet unforeseen developments in the Pakistan and the ), all of them by this information and technological revolutions. stage declared, will either have joined in An array of procedures, techniques and the process or will be ready to. Naturally, technologies that would be used to verify those nations under suspicion of having complete nuclear disarmament has been secretly produced nuclear weapons, such as identifi ed, researched and evaluated. In some Iraq and North Korea, will either have been cases they have already been implemented, verifi ably denuclearised or will be willing to especially in relation to Russia– bilateral become so.  e remaining nuclear weapons nuclear arms limitations, the nuclear test retained by any state, as zero draws near, will ban and nuclear safeguards. Continuing no longer be on alert status or deployed on research is needed to ensure that the latest or aircraft.  ese steps will be subject technological advances are incorporated to verifi cation, including on-site inspection. into verifi cation and that promising avenues Such verifi cation measures will set precedents . For some states, such as India, Israel and Pakistan, whose likely are explored.  e long-term aim should and provide experience for those charged intended targets are relatively close, be to preclude verifi cation from becoming with designing and implementing the much warheads designed for short-range systems would be considered ‘strategic’ a negotiating obstacle once the political more complicated verifi cation regime to for arms control purposes. will to achieve nuclear disarmament emerges. come.

 Verifi cation of a nuclear weapon-free world Trevor Findlay If this is the scenario, as the world appro- to the national and international verifi cation aches complete nuclear disarmament, the organisation. Such a system would be the following fi ve verifi cation tasks become clear: starting point for verifying the dismantle- ment and destruction of the weapons. (Some • verifi cation of the dismantlement and experts have suggested pooling all remaining destruction of existing declared nuclear weapons into a single site under international weapons and nuclear weapon facilities; auspices, although this may be considered • verifi cation that no undeclared nuclear a step too far by some nuclear weapon states weapons or facilities remain; and too tempting a target for a country with • verifi cation of restrictions or bans on a secret cache of remaining weapons). delivery systems; Before destruction could begin, the con- • verifi cation of non-diversion of fi ssion- tents of the bonded stores would require able materials to new nuclear weapon authentication to prove that they were not produc tion; and fake.  is would have to be done without • timely detection of research, development revealing sensitive design information, particu- and manufacture of new nuclear weapons. larly to international inspectors from non- nuclear weapon states. Research is under way in the 3 and the  into infallible authenti- Verifi cation of the dismantlement and cation techniques based on measure ment of destruction of declared weapons and emissions and other characteristic nuclear weapon facilities signatures.4 ‘Fingerprinting techniques’ can  e fi rst task of any verifi cation system for be used to determine that weapons purport- a totally nuclear weapon-free world will be edly of the same type are in fact identical in 5 to verify the dismantlement and destruction composition and manufacture. of all remaining declared weapons and After authentication, chain-of-custody weapon production and maintenance facili- procedures, like those developed for ties.  is is the easiest part of nuclear disarma- reductions under the  Intermediate- range Nuclear Forces  Treaty and the ment verifi cation. ( ) .The following is adapted from Tom It would begin with each possessor state fi rst and second Strategic Arms Reduction Milne and Henrietta Wilson, ‘Verifying Treaties () of  and  would be the transition from low levels of nuclear submitting to an international verifi cation weapons to a nuclear weapon-free organisation (the nature of which remains implemented. (Unfortunately, the new world’, VERTIC Research Report, no. 2, June 1999, p. 17 ff . to be determined) a declaration giving a Strategic Off ensive Reductions Treaty( ), signed by Russia and the  in Moscow on . For information on the nuclear detailed inventory of its remaining weapons verifi cation research programme of the and weapons-grade fi ssionable material and  May , provides no new verifi cation UK’s Atomic Weapons Establishment models, since it leaves the method and time- at Aldermaston, see Gary George and their location.  e weapons and materials Martin Ley, ‘Nuclear arms would be placed, if this had not already been table for achieving reductions entirely to the control research at the AWE’, Verifi ca- tion Yearbook 2001, VERTIC, London, done, in sealed containers with a unique discretion of each party.) In the case of com- plete nuclear disarmament, chain-of-custody 2001. tamper-proof tag and seal affi xed to each . See Oleg Bukharin and Kenneth warhead and container. Any untagged items procedures would be used not just for missiles Luongo, ‘US-Russian warhead disarm- but also for nuclear weapons themselves: antlement transparency: the status, subsequently discovered would constitute problems, and proposals’, Princeton a treaty violation.  e containers would be University/Center for Energy and Environmental Studies (PU/CEES) stored in secure identifi able locations, some- • to monitor the transport of the items to report no. 314, April 1999. times referred to as ‘bonded stores’, well away destruction/disposition sites; . Theodore B. Taylor and Lev P. Feo k- from any potential delivery systems.2 After • to verify the dismantling and destruction tistov, ‘Verifi ed elimination of nuclear warheads and disposition of contained declarations had been made, international of weapon components; and nuclear materials’, in Francesco and national on-site inspectors could monitor • to ensure that weapons-grade fi ssionable Calogero, Marvin L. Gold berger and Sergei P. Kapitsa (eds), Verifi cation: these sites permanently, supplemented by material is placed under international Monitoring Disarmament, Westview a range of sensors directly linked by satellite safeguards. Press, Boulder, CO, 1991.

 Verifi cation of a nuclear weapon-free world Trevor Findlay Trevor Findlay Verifi cation of a nuclear weapon-free world  Dismantling facilities could be built with ‘challenge’ activities will need to be handled one entrance and one exit. Warheads would carefully to avoid creating mistrust and be monitored entering the facility and correl- suspicion. According each former nuclear ated with the warhead ‘pits’ (reformed into weapon state the right to conduct a quota shapes that have no security classifi cation) of such inspections in the fi rst few years of and other components and materials as they a total ban entering into force—regardless of came out through the exit.  e pits would any actual suspicions or compelling evidence be placed under international safeguards of non-compliance—could achieve this. and removed to internationally monitored Random, unannounced inspections could “Increased storage facilities to await fi nal disposition.6 also be used. On-going verifi cation to ensure International inspectors would have no access that no new weapons or facilities are con- transparency, to the inside of the facility, where national structed will obviously need to be carried active co- personnel would carry out the dismantlement. out thereafter. Verifi cation would also be necessary to operation ensure the dismantling and decommissioning between of , production and Verifi cation of restrictions or bans on scientists and maintenance facilities. Ideally, they should delivery systems be razed to the ground. Satellite monitoring For the transition to zero to occur there will ‘managed- could verify this process.  e complete need to be limitations or outright bans on closure of nuclear research laboratories is access’ diff erent types of delivery systems, notably likely to be resisted because peaceful nuclear strategic , ballistic and cruise missiles inspections research will continue in a nuclear weapon- and nuclear-armed . If outright free world. But increased transparency, could help bans are impossible to negotiate, specifi c active co-operation between scientists and to ensure numbers of delivery systems may be per- ‘managed-access’ inspections could help to mitted for conventional weapon delivery ensure compliance. compliance” purposes, or, in the case of ballistic missiles, for space launch purposes, although such exceptions would make verifi cation more Verifi cation that no undeclared weapons diffi cult. or facilities remain A great deal of experience has been and While there will ideally be a great deal of will be further accumulated with regard to trust between states that their declarations verifying numbers of deployed strategic of remaining nuclear weapons are complete bombers and ballistic missiles. As a result, and correct—signifi cant preparatory work universal restrictions or bans on these items and experience will serve to reinforce such could be verifi ed with a high degree of trust—there may still be lingering suspicions. confi dence. Satellites and aircraft overfl ights,  ese will need to be dealt with sooner rather along the lines of the  Open Skies Treaty than later. A challenge inspection system will, (which has now entered into force) will be therefore, need to be in place from the used for such purposes. Banned strategic outset to permit the conduct of intrusive bombers, ballistic missiles and missile silos on-site inspections. Using technologies like will have to be destroyed and any suspected ground-penetrating radar, environmental deployment sites subject sampling and overhead imagery, such inspec- to ‘anytime, anywhere’ on-site inspections. tions should be able to determine that at Missile manufacturing plants will be subject least identifi able sites of concern are free of to the type of inspections seen in the  nuclear weapons or materials. Provision treaty to ensure that only missiles of the will need to be made for remote monitoring permitted type and range are being built. . Milne and Wilson, p. 21. of such sites if that is felt necessary. Such If range restrictions are imposed, the type

 Verifi cation of a nuclear weapon-free world Trevor Findlay of tests conducted in Iraq by the United weapon-free world, there will continue to Nations Monitoring, Verifi cation and be a need for nuclear safeguards to prevent Inspection Commission () could the diversion of nuclear materials from verify them. Missile test ranges could be peaceful employment to weapons. Such a verifi ed from the air. Space programmes system would be based on, but be even more and space launches will need to be monitored stringent than, the strengthened safeguards intrusively to ensure that they are not used system currently being implemented by the to mask missile development programmes International Atomic Energy Agency (). and to ensure that space launches are for A safeguards system in a nuclear weapon- peaceful purposes.  e development and free world would need to cover all nuclear deployment of anti-ballistic missile () material worldwide, including all weapons- systems will also need to be monitored to usable , whether in reactors, ensure that they cannot be used for off ensive stockpiles or extracted from dismantled purposes. Alternatives would be to ban weapons. Consequently, all of the nuclear  systems completely or to ensure that material holdings of the nuclear weapon a global  system was collectively owned states would come under safeguards and and operated. the distinction between nuclear and non- Verifying that other delivery systems are nuclear weapon states would disappear.  e not armed with nuclear weapons would be amount of material and number of facilities more problematic. In the case of submarines, requiring safeguards would thus increase intrusive on-site inspections in port could substantially, compared with today. If guarantee that they were no longer nuclear- weapons-usable materials— and armed, but overhead imagery would be highly-enriched ()—continued needed to ensure that nuclear weapons were to be permitted for peaceful purposes, not placed onboard by other means, such primarily in and research as from ships at sea. Non-strategic aircraft reactors, the verifi cation task would be much can be used to deliver nuclear weapons and greater than if nuclear reactors were permitted verifi cation in this case would be impossible. to use only low- (). Likewise, any cruise or short-range missiles In addition, if  continued to be used permitted for conventional purposes in a in naval propulsion, special arrangements nuclear weapon-free world could be relatively would need to be made to bring such material easily converted for nuclear use. Similarly, under nuclear safeguards. Safeguards should non-conventional means of delivering nuclear also be extended to uranium mining and weapons (for example, in a suitcase or in milling to ensure that all sources of new the hold of a ship) would be as impossible fi ssionable material are accounted for.7 to control and verify, as they are today. Such Other ways in which safeguards would diffi culties with controlling delivery systems have to be further strengthened include make it even more essential that there be increasing the intrusiveness of inspections, an eff ective global verifi cation system for lowering the quantities and increasing the detecting illicit nuclear warheads and the types of nuclear materials requiring declara- diversion of weapons-usable nuclear material tion and inspection, and boosting the intelli- to nuclear weapons. gence and data-handling capacities of the 8 international verifi cation organisation. Some . Currently they only begin when of the most important measures are outlined uranium is converted to ‘yellowcake’, below. a form suitable for fuel fabrication Verifi cation of non-diversion of fi ssionable or enrichment. materials to new nuclear weapon production . Adapted from Steve Fetter, ‘Verifying nuclear disarmament’, Occasional Since it is highly unlikely that all use of •  e current ‘signifi cant quantity’ of weap- Paper, no. 29, Henry L. Stimson Center, nuclear materials will be banned in a nuclear ons-usable material considered necessary Washington, DC, October 1996.

 Verifi cation of a nuclear weapon-free world Trevor Findlay Trevor Findlay Verifi cation of a nuclear weapon-free world 5 for producing a nuclear weapon (eight unlike those for the illicit production or kilo grammes of plutonium or kgs of diversion of fi ssionable materials, are compara- ) would have to be lowered to provide tively small and may be relatively easily greater reassurance. hidden. Illicit new production is unlikely • Other nuclear materials recently identi- to take place at old facilities, which will be fi ed as weapons-usable would have to be closely and comprehensively monitored, accorded their own standards. but, rather, at new, specially designed under- •  e standard for ‘timely detection’ would ground facilities or at remote locations. also have to be revised downward from Random and challenge on-site inspections, months to weeks, since former nuclear aerial monitoring through a cooperative weapon states could convert diverted open skies inspection regime, and satellite material into a fabricated weapon more imagery (from an internationally-controlled quickly than non-nuclear weapon states satellite system) may reduce the risks and to which the current standard applies. increase the costs of such activities to an • All remaining nuclear facilities, whether actual or potential violator. Yet it is diffi cult operating or decommissioned, would have to conceive of systematic verifi cation tech- to be monitored continuously and data niques to guarantee completely the detection transmitted in real-time to the verifi cation of such violations. headquarters by satellite link.  e possibility of detection may, however, • Intelligence information of the highest be enhanced through three variables that are quality would have to be available to the external to the formal verifi cation system. international verifi cation organisation. One is ‘national technical means’ (), • Permanent environmental monitoring, which refers to verifi cation and monitoring especially around nuclear and nuclear- capabilities under individual state control related facilities, would have to be under- and which include satellite moni toring, elec- taken, to detect normal as well as accidental tronic eavesdropping, information-gathering releases into the environment. and espionage.  ese will all continue and • Import–export regimes will need to be perhaps intensify in a nuclear weapon-free strengthened and universalised. world. Many states will require the additional assurance that national systems can provide Finally, the international verifi cation organi- before ratifying a nuclear disarma ment con- sation charged with ensuring a nuclear ven tion. While data from such systems may weapon-free world would need to have the be manipulated and used in a self-serving right to conduct virtually no-notice anytime, fashion or be misused politically within the anywhere inspections of any suspect site, state concerned, such possi bilities would be an even more intrusive system than that attenuated in a nuclear weapon-free world envisaged for the Organization for the by the existence of a strong multilateral Prohibition of Chemical Weapons () system with its own independent data collec- under the  Chemical Weapons Conven- tion and analysis capabilities. tion () and much closer to that operated  e second complement to the offi cial by  in Iraq. verifi cation system is ‘societal verifi cation’, which employs civil society, including non- governmental organisations (s), pro- Timely detection of research, development fessional bodies and individuals (such as and manufacture of new nuclear weapons academics, scientists and engineers), to . See Dieter Deiseroth, ‘Societal  is will be one of the most diffi cult verifi ca- monitor the activities of governments and verifi cation: wave of the future?’, 9 Verifi cation Yearbook 2000,VERTIC, tion tasks in a nuclear weapon-free world, if necessary to ‘blow-the-whistle’. A nuclear London, 2000. since the facilities required for these activities, weapons convention should make specifi c

 Verifi cation of a nuclear weapon-free world Trevor Findlay mention of, and provision for, societal verifi - the Development, Testing, Production, cation. While one could not rely completely Stockpiling, Transfer, Use and  reat of Use on such methods, they certainly add to the of Nuclear Weapons and on their Elimination complexity of the task facing any would-be has already been drafted, one version of violator. Organised societal verifi cation is which was submitted by Costa Rica to the most feasible in open societies, but even United Nations () General Assembly in closed societies or open ones with secretive .12  e regime will be elaborate, intrusive programmes have diffi culty in preventing and expensive (compared with current multi- defectors and others from leaking national lateral disarmament agreements, but not security information.  e cases of Mordechai compared with the cost of maintaining Vanunu in regard to the Israeli nuclear arsenal, nuclear arsenals). While the specifi cs of such Kamal Hussein in relation to Iraq’s biological a regime are necessarily speculative, standard weapons programme and various Russian verifi cation and compliance models for defectors and ‘whistle-blowers’ are instruc- international disarmament agreements are tive. Cheap and ready access to satellite likely to be emulated.  e following outline imagery10 and the instantaneous capabilities is based on an assumption that there would of modern communications greatly increase be a single, universal nuclear disarmament the possibilities for s to participate in convention that would supersede the  verifi cation activities. Nuclear Non-Proliferation Treaty (), the A third means of supplementing formal  Comprehensive Nuclear Test Ban Treaty verifi cation is so-called cooperative threat (), the various nuclear weapon-free zone reduction programmes.  e model here is treaties and other nuclear-related treaties.13 the Cooperative  reat Reduction Program between Russia and the . Ensuring that large parts of the former Soviet nuclear com- A Conference of States Parties plex are safely dismantled, that fi ssionable  is would comprise representatives of all material from dismantled nuclear weapons states parties. Given the importance of is accounted for and disposed of, and that nuclear disarmament to all states and the former nuclear scientists and facilities are security nightmare that a rapid breakout gainfully and peacefully employed has greatly from a nuclear disarmament regime would increased transparency and confi dence entail, membership will have to be universal. . See Bhupendra Jasani, ‘Remote 11 between the two states. One can imagine  e conference would be the treaty’s ultimate monitoring from space: the resolution revolution’, Verifi cation Yearbook a similar process involving all of the nuclear decision-making body, responsible for its 2000, VERTIC, London, 2000. possessor states and those that have come overall eff ectiveness, including compliance . See Rose Gottemoeller, ‘Beyond close to a nuclear weapon capability. arms control: how to deal with nuclear by all states parties. It would be able to weapons’, Policy Brief, no. 23, Carnegie recommend amendments to the treaty, which Endowment for International Peace, in this case would have to be binding on all Washington, DC, February 2003, and Nancy Gallagher, ‘Verifi cation and parties. It would be impossible to envisage advanced co-operative security’, a nuclear disarmament treaty with selective Verifi cation Yearbook 2002, VERTIC, Institutional components of a London, 2002. adherence to amendments by states parties. . See UN document A/C 1/52/7 verifi cation and compliance and Draft Convention in Merav Datan and Alyn Ware, Security and regime Survival: the Case for a Nuclear  e international verifi cation capabilities for Executive Council Weapons Convention, International a nuclear weapon-free world will likely be  is would be a standing body, comprising Physicians for the Prevention of Nuclear War, Washington, DC, May 1999. organised and managed by a dedicated verifi - a representative selection of states parties, . This would naturally have to be cation and compliance regime established which would be responsible for day-to-day done without damaging these existing treaties (as the CWC was negotiated by and for a nuclear disarmament convention. decision-making on the operation of the without damaging the 1925 Geneva A Model Convention on the Prohibition of treaty, particularly its verifi cation and com- Protocol).

Trevor Findlay Verifi cation of a nuclear weapon-free world 7 pliance mechanisms. Constantly alert to which would manage the verifi cation system, potential non-compliance with the accord, it and an international inspectorate that would would receive a steady stream of virtually real- be responsible for on-site inspections. A time reports from the treaty secretariat based scientifi c advisory board would also be on information from the treaty’s verifi cation indispensable. As well as a headquarters, the and monitoring system.  is would permit organisation would presumably need regional it to make judgements about compliance and offi ces and offi ces in all of the former nuclear non-compliance.  e council would also weapon states and virtual nuclear weapon have the power to demand clarifi cation from states in order to liaise closely with national any state party and an immediate on-site authorities responsible both for compliance inspection anywhere on the territory of any with the treaty and for peaceful nuclear activi- state party.  e council would ultimately ties permitted by it.  e organisation would have the power to recommend action in supersede and subsume the  and its the case of non-compliance, including by nuclear safeguards system; alternatively, the referring the matter to the  Security  itself would become the organisation Council. Finally, the Executive Council could responsible for verifying complete nuclear order that improvements or adjustments be disarmament.  is organisation would also made to the verifi cation system. absorb the Comprehensive Nuclear Test Ban All of the current nuclear weapon states Treaty Organization (), along with its (declared and non-declared) would need to International Monitoring System () and be permanent members of the Executive International Data Centre (), since the Council, as presumably would all states with detection of clandestine nuclear tests would a signifi cant ‘virtual’ nuclear weapons capa- also be an integral part of its verifi cation task. bility—that is, the ability to manufacture a nuclear device within a short period by virtue of their industrial and non-military Arrangements between former nuclear nuclear capabilities and assets.14 All of these weapon states states would need to be closely involved and In addition to international arrangements, have a strong sense of ‘ownership’ of the there are also likely to be arrangements regime, since, unlike other disarmament between pairs and groups of former nuclear agreements, the existence of only one treaty weapon states, established to give them addi- ‘holdout’ would completely defeat the pur- tional reassurance as the nuclear disarmament pose of the agreement. Hence, the council process proceeded towards zero.  ese could would be a large body, perhaps needing a have been designed to endure indefi nitely small executive sub-organ to make routine or only until the multilateral system proved decision-making more effi cient. its eff ectiveness. Such arrangements would include those for the Russia– nuclear reduction treaties from   onwards and An Organisation for the Prohibition of any similar arrangements between, for exam- Nuclear Weapons (OPNW)? ple, China and the , China and India, India Some such body would be required to estab- and Pakistan and Israel and its neighbours. lish, administer and operate the treaty’s international verifi cation and monitoring . For a comprehensive discussion system. It would be staff ed by international A strengthened Security Council of virtual nuclear capabilities, see George Paloczi-Horvath, ‘Virtual civil servants and scientifi c and technical As the likely fi nal arbiter in any compliance nuclear capabilities and deterrence experts and would be headed by the equiva- dispute (as in the case of other multilateral in a world without nuclear weapons’, VERTIC Research Report, no. 3, October lent of a director-general. It would presum- disarmament agreements), and, therefore, 1998. ably include a large technical secretariat, a vital component of any compliance system

 Verifi cation of a nuclear weapon-free world Trevor Findlay for a nuclear weapon-free world, the  powers that their relationships had improved Security Council would need to be reformed. so much as to obviate the need for nuclear It is inconceivable that the current permanent weapons, the main threat to a nuclear fi ve member states, which are all nuclear weapon-free world would be a ‘rogue state’ weapon possessors, could be permitted to that had not previously produced nuclear veto action against themselves or any other weapons.  e most worrying scenario would country that violated a nuclear weapon ban. be a ‘bolt-from-the-blue’ pre-emptive strike Furthermore, all of the current nuclear by the proverbial madman—a nuclear Hitler. weapon states, declared and non-declared,  ere are, in this case, likely to be warning and all of the other major powers, most of signs of the true nature of such a regime and “The most which are also capable of acquiring nuclear its nuclear intentions, which would probably worrying weapons, would have to be represented trigger intensifi ed scrutiny by the verifi cation permanently on the council. Consequently, system in a denuclearised world. Any weap- scenario a mixture of nuclear and non-nuclear great on(s) produced would be untested, could would be a powers would comprise the permanent not be deployed until the last minute, could ‘bolt-from- members of the council, helping to de-legiti- probably not be delivered by conventional mise nuclear weapons, although continuing means, and overt training for use would the-blue’ to refl ect the actual distribution of power have been impossible. Such a scenario is, of 15 pre-emptive in a non-egalitarian international system. course, possible today, and, in some respects, is more likely given the relative weakness of strike by the existing verifi cation regimes in the absence proverbial of a total nuclear weapons ban. In the current nuclearised world such an attack is deterred madman—a The ‘breakout’ problem by the certainty of a nuclear counter-attack. nuclear Hitler” While all of the verifi cation techniques and In a nuclear-free world it would have to be institutional arrangements described above deterred by a devastating and increasingly would aim to prevent and/or deter breakout, accurate and powerful conventional assault, it could, nonetheless, occur. While in the the credibility of which would be enhanced abstract such an event might seem cataclys- by mutual guarantees by the great powers to mic, in reality, its impact would depend on come to any state’s assistance were it to be the particular circumstances: threatened or attacked with nuclear weapons.  e possibility of an illicit nuclear weapon • whether the violator used or simply threat- being suddenly revealed and used to alter ened to use a nuclear weapon (or weapons) the course of a major conventional war and for what purpose; would be presaged by the outbreak of such • the state of readiness and deliverability of a war. Eff orts would have to be made, the weapons or purported weapons; therefore, to prevent any nuclear-capable • the existence of defences against whatever state being forced into such a corner. delivery system the violator might try to If a previously hidden nuclear weapon or use; weapons were to be used for political pur- • the relative conventional military strength poses, presumably blackmail, the existence of the violator and the rest of the interna- of the arsenal would have to be revealed, tional community combined; and or at least alluded to, thereby alerting the • the international community’s willingness international community to a major treaty to respond. violation. A ‘demonstration shot’ would have . In addition to the current the same eff ect (and, humiliatingly, might permanent fi ve one could imagine adding, for instance, Brazil, Canada, Since achievement of nuclear disarmament fail). Potential responses to such breakout Germany, Japan, India, Indonesia, would require consensus among the great scenarios encompass political, economic and Nigeria and South Africa.

Trevor Findlay Verifi cation of a nuclear weapon-free world 9 military sanctions, including the use of con- will need to be drawn into the process of ventional force. Missile defences against putting in place the essential precursors for ballistic missiles and aircraft could decrease a verifi able total nuclear disarmament treaty. the threat of nuclear blackmail and actual  e following are among the most important. attack. One element of deterrence would be the residual ability of some states to reconstitute Nuclear transparency “One element quickly a nuclear device or arsenal. For the  e sooner transparency can be achieved in former nuclear weapon states, depending relation to numbers, types and deployments of deterrence on how long a nuclear weapon-free world of nuclear weapons, delivery systems and would be the had existed, this might amount to only a holdings of special nuclear materials, the month or two. While an obviously impen- earlier and deeper can confi dence be estab- residual ability ding threat could be countered in this way, lished. Openness about past production of of some states it would be at the risk of reigniting a nuclear fi ssionable materials will be particularly . An alternative suggested by some challenging, since, even with the best inten- to reconstitute observers is a small deterrent arsenal under tions, it will be virtually impossible for any quickly a international control, although this would nuclear weapon state to give a completely nuclear device raise command-and-control diffi culties and accurate account.  e ’s experience in would be incompatible with total nuclear verifying South Africa’s account of its past or arsenal” disarmament. production, even with a high degree of co-  ese hypothetical scenarios notwithstand- operation from the South African authori- ing, neither the technology of verifi cation ties, is salutary.  e documentation of past nor the broader verifi cation and compliance production (‘nuclear archaeology’) must system can solve the breakout problem alone. begin now, while any glaring discrepancies Verifi cation can never provide complete discovered are not strategically signifi cant assurance that a small clandestine nuclear and potentially destabilising. arsenal or hidden cache of plutonium will be discovered. What verifi cation can do is to reduce signifi cantly, albeit unquantifi ably, Confi dence-building measures the likelihood of breakout occurring. It does  ese should include exchanges by the so through a combination of deterrence nuclear weapon possessors on the accepta- and enhanced warning time through early bility of various intrusive verifi cation detection. techniques and growing familiarity with each other’s nuclear establishments and facilities through exchange visits and co- operative monitoring ventures.  is process is likely to begin with France, Russia, the Necessary precursors for  and the , but needs to be quickly eff ectively verifying a nuclear extended to China, India, Israel and Pakistan. weapon-free world Such a dramatic expansion in the scope and intrusiveness of verifi cation as envisaged Deepening experience of nuclear and other above will probably require an incremental, verifi cation regimes ‘learning-by-doing’ process of increasing A key precursor of a verifi cation system for transparency and confi dence-building over nuclear disarmament will be American and many years. In addition to deep cuts by the Russian experience of verifying deep cuts, two largest nuclear weapon states, Russia and building on their already extensive bilateral the , all of the other nuclear weapon states experience in verifying the  and the

 Verifi cation of a nuclear weapon-free world Trevor Findlay   and  treaties.  e lessons need to Such a world, clearly diff erent from our be shared with all nuclear weapon states. own, but not impracticably idealistic, is Multilateral experience in verifying the likely to be arrived at gradually, through , the  Biological Weapons Conven- increasing transparency, confi dence-building, tion () and the  will also be germane, an evolving attitude towards the utility of particularly in regard to on-site inspections nuclear weapons, growing experience of and the operation of global multilateral verifi able interim steps towards nuclear . See Tom Milne, ‘Global spending on nuclear disarmament verifi cation monitoring networks. Valuable lessons have disarmament, and the progressive involve- work’, Verifi cation Matters, no. 3, already been learned, including that on-site ment of all of the nuclear weapon states— VERTIC, London, April 2002. inspections can be managed in a way that declared and undeclared. . For further information see VERTIC’s four ‘Getting to Zero’ reports: Patricia does not reveal security or commercial pro- Yet there can be no foolproof guarantee Lewis, ‘Laying the foundations for prietary information and that some of the against unexpected breakout through the getting to zero: verifying the transition to low levels of nuclear concerns that states have prior to negotiating retention of hidden stocks or the manufacture weapons’, VERTIC Research Report, intrusive regimes fall away once implemen- of new ones.  is scenario must, however, no. 1, September 1998; Tom Milne tation begins and experience grows. be seen not just in the context of the veri- and Henrietta Wilson, ‘Verifying the transition from low levels of nuclear fi cation and compliance systems established weapons to a nuclear weapon-free specifi cally for a nuclear disarmament treaty, world’, VERTIC Research Report, no. 2, June 1999; George Paloczi-Horvath, Research and development but also in the evolution of the international ‘Virtual nuclear capabilities and Currently, the vast bulk of research into system between now and then. States will deterrence in a world without nuclear have to have made signifi cant changes in weapons’, VERTIC Research Report, verifi cation procedures, techniques and no. 3, October 1998; Suzanna van technologies is conducted in the .16 Other their attitudes to the limits of sovereignty, Moyland, ‘Sustaining a verifi cation nuclear weapon states need to establish their the rule of international law and governance regime in a nuclear weapon-free world’, VERTIC Research Report, no. 4, own programmes, not only because they of the international system, particularly in June 1999. need to be convinced of the capability of regard to enforcement, for nuclear disarma- Also see: ‘Background Papers’, ment to be negotiated. Canberra Commission on the various standard verifi cation techniques, but Elimination of Nuclear Weapons, also because they could develop innovative  e attainment of a nuclear weapon-free Commonwealth of Australia, Canberra, techniques and technologies themselves.  e world is so dependent on such changes that August 1996, especially numbers 8, 11–14 and 17; Joseph Rotblat (ed.), non-nuclear weapon states should also be we will only be able to judge fully and Nuclear Weapons: The Road to Zero, encouraged to conduct such research, as they accurately its verifi ability as we become Westview Press, Boulder, CO, 1998; seriously engaged in moving towards that Robert Green, The Naked Nuclear did in the lead-up to and during negotiations Emperor, The Disarmament and on the  and the . world. In doing so, we need to ponder Security Centre, Christchurch, New whether a world with seven declared, one Zealand, 2000; Barry Schneider and William L. Dowdy (eds), Pulling undeclared and numerous potential nuclear Back from the Brink: Reducing and weapon states is safer than a denuclearised Countering Nuclear Threats, Frank world with a strong international verifi cation Cass, London, 1998; Harold Feiveson (ed.), The Nuclear Turning Point: a Conclusion system and a remote chance of nuclear Blueprint for Deep Cuts and De-Alerting breakout.17 of Nuclear Weapons, Brookings An impressive and reliable verifi cation system Institution Press, Washington, DC, can, even on the basis of current knowledge, 1999; Merav Datan and Alyn Ware, be constructed to verify with high, although Security and Survival: the Case for a Nuclear Weapons Convention, Inter- not exactly quantifi able, certainty that all national Physicians for the Prevention parties to a universal nuclear disarmament of Nuclear War, Washington, DC, May 1999; Annette Schaper, ‘Verifying treaty are complying with their obligations. nuclear arms control and disarma- Verifi cation can increase the risks of detection ment’, in Verifi cation Yearbook 2000, and consequent political costs to any poten- VERTIC, London, 2000; Edward Iff t, ‘Verifying nuclear disarmament’, in tial violator, extend the warning time to Verifi cation Yearbook 2001, VERTIC, permit responses to be mounted, and foster London, 2001; Nikolai Sokov, ‘Recent develop ments in nuclear weapons mutual trust and confi dence among the verifi cation’,Verifi cation Yearbook 2002, parties. VERTIC, London, 2002.

Trevor Findlay Verifi cation of a nuclear weapon-free world 11 Editor About this paper Trevor Findlay In this fi rst VERTIC Brief, Trevor Findlay considers the requirements for verifying complete nuclear Sub-editor disarmament. He describes the verifi cation tasks that will need to be accomplished while Richard Jones ‘getting to zero’, including those related to nuclear weapons, nuclear research and development and manufacturing capabilities, nuclear-capable delivery systems and fi ssionable material. Design and production The Brief also considers the institutional components of the verifi cation and compliance Richard Jones regime that are likely to be required, ranging from a substantial technical secretariat to a Printer strengthened United Nations Security Council. Special attention is paid to the ‘breakout’ CCP, London problem—the sudden re-emergence of a nuclear weapon state in a denuclearised world— and the possible responses. ISSN Finally, the paper considers what it terms the ‘necessary precursors’ for building an eff ective 1740-8083 verifi cation system for a nuclear weapon-free world, all of which exist in embryonic form at © VERTIC 2003 present but which must be greatly developed and enhanced if ‘getting to zero’ is to succeed.

VERTIC is the Verification Research, Training and Current funders Diana, Princess of Wales Memorial Fund, Information Centre, an independent, non-profit Ford Foundation, Joseph Rowntree Chari table Trust, UK making, non-governmental organisation. Its mission Foreign and Commonwealth Office, UK Ministry of Defence, is to promote effective and efficient verification as a W. Alton Jones Foundation and Ploughshares Fund. means of ensuring confidence in the implementation of international agreements and intra-national Board of Directors Susan Willett BA (Hons), MPhil (Chair); agreements with inter national involvement. VERTIC Gen. Sir Hugh Beach GBE KCB MC; Lee Chadwick MA; Dr Owen aims to achieve its mission through research, training, Greene; Joy Hyvarinen, LLM, LLM; Dr Bhupendra Jasani. dissemina tion of information, and interaction with the relevant political, diplomatic, technical, scientific International Verifi cation Consultants Network Richard and non-governmental communities. Butler AO (arms control and disarmament verification); Dr Roger VERTIC Clark (seismic verification); Dr Jozef Goldblat (arms control and Baird House Personnel Dr Trevor Findlay, Executive Director; disarmament agreements); Dr Edward Iff t (arms control and 15–17 St. Cross Street Dr Molly Anderson, Environment Researcher; disarmament agreements); Dr Patricia Lewis (arms control London EC1N 8UW Dr Kenneth Boutin, Senior Arms Control & Disarma- and disarm ament agreements); Peter Marshall CMG OBE ment Researcher; Ben Handley, Administrator; Marita (seismic verification); Robert Mathews (chemical and Tel +44.(0)20.7440.6960 Kivilahti, Intern; Ben Mines, Intern; John Russell, biological disarmament); Dr Colin McInnes (Northern Ireland Fax +44.(0)20.7242.3266 MSc Econ, Arms Control & Disarma ment Research decommissioning); Dr Graham Pearson (chemical and biological E-mail [email protected] Assistant; Angela Woodward BA (Hons), LL.B., disarmament); Dr Arian Pregenzer (co-operative monitoring);

Website www.vertic.org cation verifi through trust Building Legal Researcher. Dr Rosalind Reeve (environmental law).