Document Number (FOIA) /ESDN (CREST): 
Document Page Count: 
Document Release Date: 
July 30, 2014
Case Number: 
Publication Date: 
June 1, 1991
PDF icon DOC_0000624364.pdf206.18 KB
Approved for Release: 2014/07/29 000624364TITLE:AUTHOR:VOLUME:OSIand ArmsControl Monitoring(b)(3)(c)35ISSUE:Summer YEAR:1991Approved for Release: 2014/07/29 000624364 Approved for Release: 2014/07/29 000624364T U.D-VE S ININTELLIGENCE ?A collection of articles on the historical, operational, doctrinal, and theoretical aspects of intelligence.All statements of fact, opinion or analysis expressed in Studies in Intelligence are those of. the authors. They do not necessarily reflect official positions or views of the CentralIntelligence Agency or any other US Government entity, past or present. Nothing in thecontents should be construed as asserting or implying US Government endorsement of anarticle's factual statements and interpretations.Approved for Release: 2014/07/29 0006243640 Approved for Release: 2014/07/29 000624364A formidable taskOSI and ArmsControl MonitoringAnew age of arms control treaty implementa-tion has been ushered in by the on-siteinspection (OSI) associated with the Inter-mediate Nuclear Forces (INF) Treaty, the Thresh-old Test Ban Treaty (TTBT), the Treaty on Con-ventional Forces in Europe (CFE), and, most 'recenty, the Strategic Arms Reduction Treaty(START). There is, furthermore, a bilateral chemi-cal weapons agreement with the Soviets to beimplemented, and some prospect for the multilat-eral Chemical Weapons Convention to be con-cluded next year. These also contain provisions forOSI.In the early 1980s, there was little expectationwithin the US administration that the USSRwould ever accept OSI provisions. Previously, theUS had proposed OSI as early as the 1960s, whennuclear testing limitations were first being negoti-ated, resulting in the characteristic Soviet refusal.Except for the Peaceful Nuclear Explosions Treatyof 1976, the Soviets had consistently refused toagree to any on-site activity.In the late 1980s, as with so much else in theUSSR and Eastern Europe within the past fewyears, the situation suddenly changed. Moscowaccepted OSI across the board, and negotiationsbegan in earnest. To a large extent, the US wastaken by surprise. Having fully expected the USSRto maintain its historic posture and continue torefuse OSI, the US was not ready to specify itsdetailed requirements and modalities for OSI.Indeed, it did not understand all the implicationsof OSI. The Soviets were even less prepared; theylooked to the US to provide all the initial draftlanguage upon which negotiations would be based.Consequently, even though the general OSI con-cept has been around for almost 30 years, a lot of1Secret(b)(3)(n)new ground had to be broken in a short time. Thisprocess continues even today. Each new treaty hasits own peculiarities, and so for each treaty newways have to be discovered to implement OSI andto use OSI information.Two Kinds of OSIOSI refers to all activities performed by the in-specting or verifying party, say the US, in theinspected or host country, say the USSR. Thereare, however, two distinct kinds of OSI. There isinspection of declared activities and facilities, andthere is inspection of undeclared facilities, alsocalled suspect site or challenge inspection. De-clared-site inspections can occur only at placesidentified ahead of time by the treaty parties.These are the only kind of inspections allowed inthe INF and TTB treaties. The concept of chal-lenge inspection, to allow access to non-declaredsites where prohibited activities may be takingplace, is contained in the CFE treaty, and it isbeing negotiated in other treaties.Advance notice of inspection has to be given in allcases. The amount of advance notice may varyfrom one treaty to another, but the inspected partywill always know ahead of time when and wherethe inspection team will arrive, how long it willstay, and what the inspection procedures will be.Moreover, the inspected party is responsible forproviding the transportation for the inspectionteam from the point of entry to the inspection site.This gives the inspected party considerable oppor-tunity to prepare the site for inspection, and, inextreme cases, to delay or prevent arrival of theinspection team. Of course, such actions would be"cause for concern." They would be discussed inApproved for Release: 2014/07/29 000624364Secret  (b)(3)(n) Approved for Release: 2014/07/29 000624364Secret(b)(3)(n)the consultative body associated with each treaty,and they could possibly result in a demarche. Butthey can be attributed to bad weather, faultyequipment, or human error, and there can be nodoubt that the inevitable protest would be viewedby the inspected party as preferable to the "smok-ing gun" detection of an actual treaty violation.OSI is expensive, and so it is fair to ask thefundamental question, "What good is OSI?" Anargument has been made that there is little likeli-hood of a treaty party trying to cheat at aninspection site while inspectors are present. If onebelieves this statement, it follows that there is littlethat OSI can do to help catch cheating, becausesuch cheating presumably would be taking placesomewhere or at some time when inspectors arenot present.OSI is often called a "confidence-building meas-ure" (CBM), meaning that if the inspections dem-onstrate that one side is accurately providing in-formation about the inspected sites, then the otherside has some increased confidence that the provi-sions of the treaty are being observed. CBMsprobably provide some political benefit.Leaving political value aside, the intelligence ques-tion to be asked is, "How much, if at all, does OSIimprove treaty monitoring?" Some qualitative val-ue for National Technical Means (NTM) can beexpected from OSI in all treaties from the addedinformation that may help us better use the datafrom intelligence sources. In only the TTBT treaty,however, can a quantitative improvement of NTMbe readily determined. It may be that only theTTBT offers the opportunity for suchimprovement.The TTBTIn the TTBT and its associated Joint VerificationExperiment (JVE), the inspectors have the oppor-tunity to acquire fundamental data about thegeology at the nuclear test site and also to measurethe yield of the nuclear explosion more accuratelythan ever before. Besides providing direct yieldestimates, these data tell quite a bit about how the (b)(3)(n)Monitoringobserved explosion, or any other past or futureexplosion in the same vicinity, interacts with theEarth. In other words, the on-site data tell us aboutthe seismic disturbances caused by the explosion,help us to understand better the seismic signature,and consequently have a direct impact on thequality of NTM monitoring. The single data pointobtained from the JVE has already allowed signifi-cant improvement. As more such data is gatheredover time, further improvement will be possible.(b)(1)(b)(3)(n)2Approved for Release: 2014/07/29 000624364 Approved for Release: 2014/07/29 000624364Monitoring(b)(1)(b)(3)(n)The Joint Verification ExperimentThe JVE was an outgrowth of the Nuclear TestingTalks. It was conducted during the summer of1988, and it served as a rehearsal for actualverification activities which will take place underthe provisions of the two treaties limiting under-ground nuclear tests to 150 kilotons. In the JVE,the US and the USSR each measured the yield of anuclear weapon test near 150 kilotons conductedby the other country.According to the JVE agreement, the JVE wasconducted for the "purposes of elaboration ofeffective verification measures for the Treaty Be-tween the United States of America and the Unionof Soviet Socialist Republics on the Limitation ofUnderground Nuclear Weapon Tests of 1974(TTBT)." The JVE consisted of two nuclear explo-sions, one at the US Nevada Test Site (NTS) andthe other at the USSR Semipalatinsk Test Site(STS), each to have a planned yield between 100and 150 kilotons.In addition to teleseismic measurements, each sidehad the opportunity to measure and record localgeological and geophysical data. Each side also canconducted on-site measurements of the yield ofeach JVE explosion using a scheme called hydro-dynamic yield measurement, which, being done atthe site of the nuclear explosion, is inherentlymuch more accurate than the seismic technique.For the Soviet JVE, the initial on-site hydrody-namic yield estimate was 118 kilotons, with an3Secret  _(b)(3)(n)uncertainty (F-factor) of 1.16. This means thatthere is a 95-percent certainty that the actual yieldof thq exillncinn wac hpturpen 1 10 a nil 11'7 kat-,  tons.  (b)(1)  (b)(3)(n)Another way to express the quality of these esti-mates is in terms of the probability that the actualyield was above the treaty-obligated threshold of150 kilotons. From the OSI estimate, there is avery low probability?essentially zero?that the  JVE was a violati(b)(1)  (b)(3)(n)For the pur-pose of determining if the test was under 150kilotons, as required by the TTBT, the on-sitemeasurement is preferable to the teleseismic.The JVE provided the US with the first validatedyield of a Soviet nuclear weapon test. As such, it isa unique and invaluable addition to the informa-tion on which we base our NTM estimates ofyields of Soviet nuclear weapons tests. It wasstated in the JVE agreement that"Because the JVE is not designed to producestatistically significant results, it cannot byitself establish statistical proof of the accura-cy of any particular yield measurementmethod."Although the US insisted on this statement, it hasno real practical significance. The phrase "byitself" is key in the following discussion, becausethe JVE result was not used by itself, but ratherwas combined with an extensive body of seismicdata to improve?decrease?the uncertainty inour estimates of Soviet underground nuclear tests.(b)(1)(b)(3)(n)Approved for Release: 2014/07/29 000624364Secret  (b)(3)(n) Approved for Release: 2014/07/29 000624364Secret  (b)(3)(n)(b)(1)(b)(3)(n)OSI's Treaty-Monitoring ValueIn only the TTBT is it possible to use informationgathered during OSI to improve directly our meth-od of monitoring Soviet activities when inspectorsare not present. Only TTBT OSI offers the pros-pects of quantitatively improving treaty monitor-ing via NTM.Secret7(b)(3)(n)MonitoringThe single JVE has already had a profound effecton the quality of teleseismic yield estimates bysignificantly reducing uncertainties of yield esti-mates for explosions at the Soviet JVE site atShagan River. US NTM now has a much bettercapability for estimating the yield of Soviet nucle-ar tests at the Shagan River area. As on-site yieldmeasurements are made at other locations, UScapabilities using NTM also will improve for theseareas.Given a choice, on-site yield measurements wouldalways be preferable. At its best, a teleseismicestimate does not provide the confidence possiblefrom OSI. But when OSI is not available, forwhatever reason, the US still has an improvedNTM method because of past OSI opportunities.This article is classified SECRET4Approved for Release: 2014/07/29 000624364_(b)(3)(n)_