VERIF FYING A B BAN ON NU UCLEAR T TEST EXP PLOSIONS S Multilat teral Verific cation, Coll lective Secu urity: The C Contribution n of the CTB BT Vienna Internation nal Centre, A Austria, 18 J July 2012 By And dreas Persbo o, Executive e Director Let me first thank t the CTBTO O team for g getting me h here on time e. It is of cou urse fantast tic to be bac ck in the V Vienna Inter rnational Ce entre. And, of course, i t is good to see so man ny of you he ere today. Y You have ha ad a full age enda, and yo ou have the best before e you with to omorrow’s simulation of the Executi ive Council . Now, I have been a asked to say y a few wor rds about th e effective v verification n of the Com mprehensive e Nuclear r-Test-Ban Treaty. The ere is a lot to o say about that, especi ially with th he new repo ort by the Nationa al Academy y of Science es underpinn ning the deb bate. Alongsi ide the Aca ademy repor rt, I’d like to o recommen nd a book b y Ola Dahlm man, Jenife er Mackby, Svein M Mykkeltveit and Hein H Haak. It is c alled ‘ Detec ct and Dete er: Can Cou untries Verif fy the Nucle ear Test Ba an? ’ and it g goes through h the techni ical capabili ities of the C CTBT’s ver rification sy ystem in gre eat detail. S So this year r has seen a number of v very author ritative stud ies on the tr reaty. But w what do they y all say? When th he CTBT w was opened f for ratificati tion more th han 15 years s ago, the en nvisioned v erification system was already y more soph histicated th han all previ ious regime es. It was the he first treaty y ever to ha ave a regim me that relied d primarily on remote m monitoring techniques rather than on-site insp pections. It was th he first syst tem that tran nsmitted un nprocessed d data directly y to the state te parties, en nabling them m to draw th heir own con nclusions on n complianc ce. It also p laced the D irector-Gen neral and the e Secretaria at in a rather r passive rol le. Tibor To oth has some etimes refer rred to this as the ‘dem mocratization n of verifica ation’. In addit tion, it was the first sys stem that rel lied on a lar rge and sop histicated g global comm munications infrastru ucture, relay ying data fr rom the sour rce event to o the treaty o organization n and its sta ate parties in n near-rea al-time. At the t time, it all fe felt a bit like e science fic ction. But a s you now k know, scien nce has now w caught up with fic ction. And in n some case es, even sur rpassed it. Twenty y years ago, the system was envisio oned as bein ng able to d detect any nu nuclear explo osion larger r than on ne kiloton. L Let us put th hat in perspe ective. One kiloton is o one million k kilograms o of TNT. Or, , put a differe ent way, ab bout 4,000 ra ather large p precision-gu uided bomb bs going off f in one tiny y space in ab bout a micro osecond. Th at is a very large explo osion. Twenty y years later r, today, the system is f far more cap pable—dete ecting man-m made seism mic events in n the rang ge of tens of f tons in cer rtain region ns of the wor rld. Yet som me critics, e especially in n the United d
States, c continue to argue that t the treaty is s unverifiab le or that it cannot prov vide 100 pe ercent confide ence that oth her states ar re complyin ng with the n nuclear-test ban. There c could be a n number of po ossible expl lanations fo or this. Som e may belie eve that eve n very smal ll nuclear -test explos sions, not m more than ten ns of tons, a are militarily y significan nt. Others m might argue t that any inst tance of non n-complianc ce, no matte er how sma all, would se eriously und dermine the treaty and, by extensio on, the natio onal securit ty of the stat te parties. S Still others a argue that th heir country y must retain n the opti ion to condu uct nuclear tests and th hat the test b ban therefor e constrains s future cho oices. Many y of these ar rguments, b but not all, c can be count tered. No ve erification r regime can p provide one e hundred p per cent ass surance, and d no sane pe erson would d expect it t to. The IM MS, by the w way, does no ot exist in sp plendid isol ation. It can n be enhanc ced, and in t he future I think it will be. Th he IMS is far r more sens sitive when combined w with states’ national tec chnical mea ans of verif fication than n it is by itse elf. For exa ample, there e are now m ore than 16 6,000 seismi ic stations around the world, o organized in n several ne etworks. Tw wo continent ts, Europe a and North A America, are e virtually y covered w with them. In n addition, the ‘ring of f fire’ aroun nd the Pacifi fic Ocean is lined with hundred ds upon hun ndreds of sta ations. Thes se could be used for tes st ban treaty y verificatio on also, signific cantly increa asing the se nsitivity of the regime as a whole. . Add to that ever-in ncreasing sa atellite capab bilities and readily-ava ailable high- -resolution commercia al imagery y, and it bec comes clear r that it is in ncreasingly d difficult to h hide not onl nly a nuclear r test, but al lso nuclear -test site pre eparations. The authors s of Detect and Deter c call this ‘pre ecision mon nitoring’ an nd describe e it as focus sing state as ssets on one e particular region of th he world. The scie ence is prog gressing als o. Seismolo ogy has bee n equipped with new m methodology y in recent years, w which shoul ld enable sta ates to bette er understan nd and use d data obtaine ed both in cl lose proxim mity to or at some distan nce from an n event. The sys stem is very y sensitive e even withou ut national te echnical me eans. The de etection thre eshold can b be as low a as magnitud de 2.8 at the e 10 per cen nt confidenc ce level. Thi is is a very s small event t. In some cases, m magnitude 2 2.8 correspo onds to a yie eld of less th han 10 tons of kilogram ms of TNT. Those 4,00 00 smart b ombs I’ve t talked about t earlier hav ve been redu uced to only y 40. When th he IMS was s initially de esigned, the e auxiliary s stations wer re envisione ed as provid ding on-call data to t the system, should the primary ne etwork detec ct a suspicio ous event. It t was consid dered costly y and unn necessary to o bring these e stations on n-line. Of c course, the d designers co ould not hav ve foreseen a rapid de ecline in the e price of da ata transmis ssion or, for r that matter r, the conseq quences of the evolutio on of silico on memory chips. The last decade e has seen a radical tran nsformation n of our virtu ual world. In the 1 990s, the C CTBTO’s ba andwidth re equirement w was seen as staggering . Today, its data can be e easily c carried on th he internet. M Moreover, m many auxili iary stations s are able to o provide ne ear-real-tim me data to t the IMS, in ncreasing the e sensitivity y and respon nsiveness o f the system m. And the r rapid evolut tion of comp puting powe er and trans smission spe eeds shows few signs o of slowing d down. So wha at does this m mean? Let m me give you u the words of John Wa alker, anoth her CTBT v eteran. He said, in a recent VE ERTIC brie f on verific ation and d eterrence, a and I quote: ‘A regime that can demons strate a very y high level of technica al reliability y, coverage a and sensitiv vity presents s a formidab ble obstacle e to anyone who wants s to cheat. T The IMS doe es that.’ End d of quote. I I agree.
Recommend
More recommend
