Silencing the Bomb Page 13

600–800 kt and larger “In U.S. experience an mb greater then 6.2 (as measured for the largest Soviet events) has only been seen for yields of 600–800 kt and larger.”

  JUDITH MILLER—NEW YORK TIMES, JULY 26, 1982

  300 kt “One official said that there had been several Soviet tests, many at one particular site, that had been estimated at 300 kilotons.”

  Source: Sykes and colleagues, 1983.

  Why did the United States start testing near 150 kilotons soon after the TTBT became effective, but the Soviet Union did not? The United States knew from previous tests that it could detonate an explosion of 150 kilotons in Nevada without any damage occurring in the nearby cities of Las Vegas and Reno. The Soviets had conducted larger tests than magnitude 6.2 before the TTBT became effective in 1976 at their remote Arctic test site at Novaya Zemlya. If they had been set off in Eastern Kazakhstan, their stronger shaking likely would have caused damage in the nearby city of Semipalatinsk. This may explain the gradual increase in yields over three years (figure 10.2) to ascertain how much damage larger tests would cause in that city.

  Soviet explosions larger than about magnitude 6.2 or 150 kilotons might well have produced damage in Semipalatinsk. In 1984 Evernden and I claimed that if a 75-kiloton explosion were set off at the site of the 1964 Salmon explosion in Mississippi, it would have caused damage to nearby towns and cities. Our reasoning was that propagation of seismic waves would be as damaging in Kazakhstan as it would be in Mississippi (or the rest of the central and eastern United States). A recent example of efficient wave propagation and strong shaking can be seen in the damage to the Washington Monument and the National Cathedral in Washington, DC, which was 100 miles (160 km) away from the magnitude 5.8 earthquake near Mineral, Virginia, in August 2011. Soviet explosions in Eastern Kazakhstan of 75 kilotons generated seismic waves of about mb 6.

  At the 1983 symposium, Inés Cifuentes, a graduate student at Lamont, and I presented new work on the estimation of Soviet yields at the Eastern Kazakhstan test site from 1978 through 1982 using digital recordings of seismic surface waves. Our work, published in 1984, was prompted by a letter I received on December 15, 1982, from Peter Marshall, who had long worked on seismic verification in Britain, about my use of surface waves for yield determination in our 1982 Scientific American paper. Marshall wrote, “I have always thought that [the surface wave magnitude] Ms should be a stable indicator of yield. Some LR [long-period Rayleigh wave] trains from the RTS [Russian Test Site in Eastern Kazakhstan] have shaken my confidence in that there are examples of where the average Ms has been significantly reduced in amplitude in all azimuths [directions].”

  What concerned Marshall and others, understandably, was that some explosions at that site triggered the release of large amounts of natural stress in the surrounding rocks. The tectonic release at the time of explosions in hard rock, depending on the amounts of natural compressive stress built up before hand, reduces the size of the surface waves from shots in Eastern Kazakhstan. Because those surface waves are reduced in amplitude, the yields computed from them from Ms are too small. Tectonic stresses in Nevada, however, are extensional, not compressive, and have the opposite effect on the size of surface waves.

  Cifuentes and I dealt with tectonic release in a study of twenty underground explosions at Eastern Kazakhstan. We used the ratio of the amplitudes of two surface waves—Love waves/Raleigh waves—as a measure of the amount of tectonic release that was contaminating Raleigh waves produced by the explosions themselves. Love waves are generated only by natural tectonic stress release, not the explosion itself. We calibrated a new mb-yield relationship for Eastern Kazakhstan by calculating yields from surface waves, but only for those explosions that were characterized by a small-to-negligible component of tectonic stress. We then used that relationship to determine yields from mb for all twenty explosions.

  We concluded, “The yields of the seven largest Soviet explosions are nearly identical and are close to 150 kilotons, the limit set by the Threshold Treaty.” For explosions characterized by a large Love to Raleigh wave ratio—that is, a large tectonic component—we recommended using just our new mb-yield relationship to determine their yields from body waves (mb) alone. This is what I had hoped and expected that AFTAC would do in 1977, but they did not.

  Our position was slowly becoming clear to journalists. John Wilke wrote about the AGU symposium in the Washington Post on June 3, 1983. He mentioned a classified study at the Livermore Lab, stating that a Livermore physicist said publicly to him, “Whatever it is believed in Washington, it is now clear that officials here at Livermore Lab do not believe that the Soviets have violated the 150-kiloton limit.”

  Journalist R. Jeffrey Smith, who also covered the symposium, quoted seismologist Bernard Minster of UC San Diego in the June 17, 1983, issue of Science: “Based on what I heard this morning, I think we have a hard time justifying statements that the Soviets are cheating.” Robert North, another consultant to the Defense Department, said, “After listening to the presentations…most people would agree that you cannot assert that the Soviets have violated the Treaty.” Smith went on to state, “Milo Nordyke, who directs the verification program at Lawrence Livermore National Laboratory, where the bulk of the government’s analysis is conducted, said after the AGU meeting ‘DARPA, which takes the most conservative view, certainly seems to be in the minority. Of course, with the conservative view, you automatically get some evidence of Soviet violations. But you have to use the best estimate, not the most conservative one. This is a message that the politicians in Washington have a hard time understanding.’” Pike and Rich of the Federation of American Scientists also quoted Nordyke as saying “there is no hard evidence of Soviet test violations.”

  DARPA and AFTAC supported several analyses of the attenuation of short-period P waves beneath test sites in Eastern Kazakhstan, Nevada, Mississippi, Amchitka Island in the Aleutians, and Algeria. In January 1984 Robert Blandford and colleagues at the consulting firm Teledyne Geotech computed corrections to mb magnitude for each site with respect to Nevada. They reported that an mb of 6.17 was associated with 150-kiloton explosions at Eastern Kazakhstan, very similar to the results we presented in 1983. They found that P waves from French explosions at the Algerian hot spot were attenuated by about the same amount as those in Nevada and the attenuation of P waves beneath Amchitka was similar to those beneath Eastern Kazakhstan.

  DARPA PANEL MEETINGS ON YIELD DETERMINATION

  In January 1983, I was invited to be a member of a Department of Defense Technical Review Panel on Threshold Test Ban Treaty Verification Issues that was convened by DARPA. After attending its first meeting, I informed DARPA that I would be out of the country and not able to attend the next meeting set for about July 29. I stated, “I have had a fairly long telephone conversation with Gene Herrin, and will send him written comments in response to his unclassified letter. I also think that it is important that my views be represented along with those of Sean Solomon [of MIT]. I understood that Sean also will not be able to attend.”

  On August 17, 1983, Alewine, who had become director of the Geophysical Sciences Division of DARPA, sent me a letter informing me that a separate classified package of the completed panel report had been mailed to me. His unclassified letter states, “As we discussed in the Panel meetings, the methodology for the use of surface waves in yield estimation has not received the level of critical review as has that for body waves.” He wrote, “We would like to convene the Panel about mid-October to examine fully the status of using surface waves for TTBT monitoring…we would like for the Panel to begin review of technical aspects of Comprehensive Test Ban Issues at the October Panel meeting.”

  I wrote an unclassified letter to Herrin on September 16, 1983, with my general views about the Panel Report and a classified letter to DARPA about some specific points. I stated, “I conclude that the panel’s recommended value for the bias [in magnitude between the Nevada and Eastern Kazakhstan] is still too low. A consequenc
e of this, of course, is that calculated yields will still be too high.” On May 31, 1983, Under Secretary of Defense Richard D. DeLaure wrote to Frank Press, then the president of the National Academy of Sciences, stating that he foresaw no need for a parallel review effort on questions of verifying nuclear test bans to be conducted by DARPA and that all legitimate questions of objectivity and credibility were well met by the [present] DARPA panel.”

  I also stated in my letter, “I am seriously concerned about the procedures that have been followed in previous panels of which I was a member. Let me be specific on three points. DARPA officials, Bache, Alewine, were present at all of the sessions of the panel at its January 1983 meeting. They have consistently and often reiterated their views that the bias between those two test sites is small. They have actively and aggressively participated in the deliberations of the panel as if they were members. Under those circumstances, I believe that it is difficult for different views to be heard and to be considered in a thoughtful manner.

  “Secondly, the panel has adopted positions that are at odds with those of several distinguished scientists including Dr. Peter Marshall, of the United Kingdom, Springer, Rodean and [Peter] Moulthrop of Livermore, Evernden, etc. I believe that none of them has had a chance to respond to criticisms of their work made in either the latest or earlier reports of the panel. Thirdly,…my sense is that the panel has had too little time during its meetings to write a thoughtful and independent report of its own.”

  I sent copies of my letter to Alewine and the panel members. I was not invited to the October 1983 meeting. I asked panel member Tom Jordan of the University of Southern California if I was still on the panel, and he said no. I never received a letter from DARPA to that effect. I was not invited to be a member of future DARPA or AFTAC panel meetings.

  The DARPA panel continued to meet until at least 1985, because I received an unclassified draft summary of their work dated February 1985. It discussed the contamination of surface waves from tests in Eastern Kazakhstan and recommended not using them for yield determination. They did not acknowledge the significant problems of using just P waves for yield determination. I was told that members of the panel disagreed about magnitude bias and wanted sections put in about it, but that material was removed from the final report.

  Writing in the New York Times on April 2, 1986, Michael Gordon, the military affairs correspondent, stated, “The Central Intelligence Agency has changed its procedures for estimating the yields of large Soviet nuclear tests because it has decided its previous estimates were too high, Reagan Administration officials said today…. Experts familiar with the change said it would lower estimates of the yield of Soviet tests by about 20 percent.”

  Gordon went on to say, “On Oct. 18, a panel of scientists selected by the Defense Advanced Research Projects Agency prepared a classified report that concluded the Government’s method for estimating the yield of Soviet explosions was based on faulty assumptions. The panel’s report was submitted in late October to the Joint Atomic Energy Intelligence Committee, which issues reports on the size of foreign nuclear explosions. The committee is made up of members from the military services and intelligence agencies…. On Dec. 17, the Joint Atomic Energy Intelligence Committee recommended that the C.I.A. adopt the advice in the report commissioned by the research agency. Officials said the Defense Intelligence Agency disagreed, but was overruled.

  “Officials said applying the new method retroactively would still leave about a dozen Soviet tests that appear to be above the limit, and one official said only three or four of these exceeded the limit enough to warrant special concern.” These statements imply that the bias in magnitudes for Eastern Kazakhstan was about 0.2, not 0.3 to 0.4 as several of us had recommended. I had assumed, apparently incorrectly, that the U.S. government had corrected the formula for estimating Soviet yields in 1977. Gordon’s article indicated that it did not occur until 1986. U.S. formulas for estimating Soviet yields have changed with time and are still classified.

  Gordon also stated, “Richard N. Perle, the Assistant Secretary of Defense for International Security Policy, reportedly opposed adopting the recommendations and argued that the issue needed more study. Mr. Perle declined to discuss the issue. Administration experts, who asked not to be identified, were divided about whether the change should lead the Administration to drop its allegations against the Soviet Union.”

  PERLE AND THE SCIENTISTS

  Brian McTigue produced a television interview in 1986 for San Francisco station KRON called “Richard Perle and the Scientists: The Controversy Over Nuclear Testing.” They filmed it during the debate about alleged Soviet cheating on the Threshold Treaty. It focused on comments made by seismologists Charles Archambeau of the University of Colorado, James Hannon of Livermore, and me, countering statements by Perle. Perle served in the Reagan administration from 1981 to 1987 as assistant secretary of defense for international security policy and later as chairman of the Defense Policy Board, an influential group of advisers to the Pentagon. The program can be viewed on youtube.com.

  I single out Perle here because he was a hardliner and the strongest opponent of the control of nuclear weapons for more than fifty years. He exaggerated Soviet nuclear capabilities on many occasions. What is particularly relevant here is that he knew nothing about seismology, which was at the very heart of estimating Soviet nuclear yields and possible cheating. Congressional, arms control, and scientific communities knew him as the “Prince of Darkness.”

  Here are excerpts from the 1986 interview.

  INTRODUCTION: The Soviet Union has stopped testing and is abiding by a nuclear test ban, but the United States has continued to test weapons underground claiming the Soviets have cheated. The Administration says the Soviets tested weapons more powerful than agreed to in 1974, but have they?

  Rollin Post and our Target 4 Investigative Unit found the Administration is ignoring evidence that the Soviets never cheated, that they are following the treaty.

  CHARLES ARCHAMBEAU [SEISMOLOGIST, COLORADO]: If the scientific data doesn’t quite agree with your political position, what is done is to bend the data a little bit.

  PERLE: Baloney! It is not a question of scientific evidence. It is a question of scientists playing politics. I’ve looked carefully at the evidence and have concluded as President Reagan did that there is significant evidence that the Soviets have violated the 150-kiloton threshold.

  ROLLIN POST [INTERVIEWER]: This has been the position of the Administration since 1983. But it has also caused a rebellion among the very scientists the Defense Department relies on to estimate the size of the Soviet tests. Target 4 interviewed some of those seismologists and they all said the Soviet Union has not violated the 1974 Test Ban Treaty.

  ARCHAMBEAU: At present there is no evidence that the Soviets have tested over 150 kilotons, none whatsoever.

  LYNN SYKES [SEISMOLOGIST, NEW YORK]: The treaty itself states that neither country should test above 150 kilotons and I have no evidence that indicates to me that the Soviets have done that.

  WILLARD HANNON [SEISMOLOGIST, LIVERMORE NATIONAL LABORATORY]: I don’t believe that the evidence supports a militarily significant violation.

  PERLE: The best experts available spent years studying this and came to the conclusion that it was likely the Soviets had violated the 150-kiloton threshold.

  Well, with all due respect he [Archambeau] is wrong, there is lots of evidence. He may not be persuaded on the basis of the evidence, but to say that there is no evidence is just flatly wrong.

  POST: Lynn Sykes of Columbia says that the Soviet Union has not violated the 150-kiloton limit of the threshold treaty as alleged. He’s wrong?

  PERLE: He’s entitled to his opinion. He’s a professor sitting up at Columbia.

  Well, all that seismology enables you is to make an estimate as to the yield of an event. Even by that standard alone, there is evidence that suggests the Soviets have violated it, your experts not withstanding. There is other e
vidence as well that is sensitive and of a classified nature.

  POST: Mr. Perle would only say the other evidence involved satellite and electronic surveillance. Target 4’s investigation learned that Mr. Perle had already convened a panel of experts, which looked at this other evidence and rejected it.

  PERLE: They came to the conclusion that of the many ways of estimating yield, seismology was the single most important and I happen to agree with that.

  I did not particularly care much what their answer was, it did not have any profound bearing on our policy.

  POST: Target 4’s investigation has also uncovered evidence that Mr. Perle improperly tried to manipulate intelligence agencies in a biased direction. Example: Perle’s letter to the Air Force when its intelligence unit asks seismologists to advise on Soviet nuclear tests. According to sources who have seen the letter, it said the intelligence community is undermining the Administration’s position. My Department will control this area. I asked Perle about the letter.

  PERLE: I don’t remember the exact words of the letter, but my concern was the concern that I have been expressing to you throughout this interview, which is that we have tended, I think wrongly, to exclude the nonseismic evidence that bears on the estimation of the yield of Soviet tests.

  They’re all seismologists; they’re a bunch of seismologists feathering their own nests. Well, seismologists have dominated this field from the beginning. It’s how they make their living. The day that it is concluded that we can get along without attributing the importance to seismology that we do—some of these fellows are going to be looking for jobs.

  ARCHAMBEAU: The scientific opinion is close I’d say to unanimous. Right now Mr. Perle finds it extremely difficult to find any scientists that will defend the DoD Perle position and that’s because there just aren’t any that believe it.

  SYKES: I think that one view that is often put forth is that arms control agreements are not in the best interest of the United States. That the Soviets will cheat and then attempt to have a self-fulfilling prophecy by coming up with some procedure, an incorrect one, that indicates to them that the Soviets have cheated.