Jeffrey LewisSeismology, Schmeizology

Seismologist Paul Richards teaches a course on Weapons of Mass Destruction at Columbia, during which he shows this amazing interview with a 1980s Richard Perle.

Perle—then Assistant Secretary of Defense and well on his way toward earning the moniker “Prince of Darkness”—is asserting the existence of “significant evidence that the Soviets have violated the 150 kiloton threshold” in the Threshold Test Ban Treaty, despite utter lack of such evidence.

The only clarification worth adding is DARPA and Perle did manage to dredge up two seismologists—Ralph Alewine and Thomas Bache—willing to claim that, based on seismic signals, the Soviets had tested a device in excess of 150 kilotons.

I recommend Alewine and Bache’s article in EOS: Transactions of the American Geophysical Union 64:193 (1983), coupled with the complete demolition of their arguments by Lynn Sykes and Ines Cifuentes in the Proceeding of the National Academies. The debate, more or less, turned on the equation for estimating yield from seismic signals at the Soviet test site.

A more readable account is the exchange of letters in Science between Alewin and Bache, on the one hand, and Sykes and Evernden that resulted from a symposium organized by Sykes, Archambeau and Everenden—all of whom appear in the KRON interview with Perle. [Full text in the comments.]

The great thing about this debate, by the way, is that we can now definitively say who was right and who was wrong. In 1988, the United States and the Soviet Union agreed to a “Joint Verification Experiment” or JVE where each side off a nuclear weapon in order to calibrate their respective test sites.

Analysis of the JVE, Sykes wrote with Goran Ekstrom* “substantiates previous conclusions about the sizes of past Soviet weapons tests and compliance with the Threshold Test Ban Treaty.”

Sadly, the Threshold Test Ban Fiasco didn’t keep Perle off the rubber chicken circuit or undercut his credibility when he was peddling later, more consequential sets of falsehoods and exaggerations.

The whole sordid story is laid out in exquisite detail by Gregory Van Der Vink in “The Role of Seismologists in Debates over the Comprehensive Test Ban Treaty,” Annals of the New York Academy of Sciences 866: 84-113.


  1. Jeffrey Lewis (History)

    Scientists Challenge Claims of Soviet Treaty Violations

    The Reagan Administration’s judgment that the Soviet Union has violated the Threshold Test Ban Treaty came under renewed attack at a recent meeting of the American Geophysical Union. A number of prominent seismologists, including Lynn Sykes of Columbia University and Charles Archambeau of the University of Colorado, presented studies to support their conclusion that the Soviets have not exceeded the 150-kiloton nuclear testing limit. “The Russians are essentially in compliance, as far as we can tell,” Archambeau said.

    Although no formal charge of cheating has been lodged with the Soviets, President Reagan has stated that “we have reason to believe that there have been numerous violations” of the treaty (Science,13 May, p. 695). He might have changed his mind after hearing the speakers in Baltimore on 2 June.

    “Based on what I heard this morning, I think we have a hard time justifying statements that the Soviets are cheating,” said Bernard Minster, a Defense Department seismological consultant.

    “After listening to the presentations . . . most people would agree that you cannot assert that the Soviets have violated the treaty,” said Robert North, another consultant.

    Support for Reagan’s claim was expressed by Ralph Alewine, director of the geophysical sciences division at the Defense Advanced Research Projects Agency (DARPA), and Thomas Bache, a program manager who works for Alewine. They said that the Soviets have detonated nine bombs above the 150-kiloton threshold since the treaty was signed in 1976.

    But Alewine conceded that six of these apparently fall between 150 and 200 kilotons and are therefore not of serious concern. “They could be due to miscalculations i n either side,” he said. Two apparently fall between 200 and 250 kilotons, and one is apparently just over 300 kilotons.

    Alewine described the latter detonations as being “egregiously above the limit.” But he noted that the Soviets would be well withirl their rights “if they just said sorry we made a mistake.” A provision of the treaty permits one or two accidental breaches of the threshold per year. Eugene Herrin, a seismologist at Southern Methodist University who chairs a DARPA advisory panel, has argued that all of the apparent Soviet violations could be due to such accidents, U.S. miscalculations, or both.

    In any event, similar accidents occur at the U.S. test site, albeit less frequently. Alewine and Bache reported that in the same period two tests in Nevada have exceeded the treaty limit by 33 and 75 percent, respectively.

    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 that “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.”-R. JEFFREYSMITH

    Nuclear Test Yields

    In his briefing about the American Geophysical Union session on the Threshold Test Ban Treaty (News and Comment, 17 June, p. 1254), R. Jeffrey Smith misquotes us as saying that, since 1976, two U.S. nuclear explosion tests have exceeded the 150-kiloton limit of the treaty “by 33 and 75 percent, respectively.” We certainly did not say this.

    On the contrary, we were careful to point out that very accurate radiochemical measurements show that no U.S. test has exceeded 150 kilotons since 31 March 1976. What we did say is that there is imprecision in the yield estimates based on seismic signals recorded at great distances and that occasionally 150-kiloton explosions will produce signals larger than expected for that yield.

    Recognizing this, we compared the largest 40 U.S. and Soviet explosions with magnitudes determined by the U.S. Geological Survey and found that only two U.S. events, but nine Soviet events, had signals larger than expected for 150 kilotons.

    This asymmetry raises serious concern that many of these tests have actual yields well over 150 kilotons. Our analysis included a fairly large adjustment (reducing the Soviet yield estimates) to correct for suspected geophysical differences between the test sites. To give some perspective on how large this adjustment must be for the Soviet yields to be less than 150 kilotons, we point out that the largest Soviet explosions produce signals that, in U.S. experience, have only been seen for yields of more than 600 kilotons and usually more than 800 kilotons. An adjustment this large is inconsistent with the best available geophysical evidence. Another important point conceded by nearly all involved is that the yields of the larger Soviet tests increased abruptly by about a factor of 2 in recent years. Thus, those concluding that the Soviets have not exceeded the 150-kiloton nuclear testing limit are also saying, by implication, that the Soviets did not test above 75 kilotons or so for the first several years of the treaty (when the United States was testing up to 150 kilotons). Why would they stay so far below an agreed limit? The question is not one we can answer but certainly is added cause for concern.

    RALPH W. ALEWINETHOMAS C. BACHEDefense Advanced Research Projects Agency, 1400 Wilson Boulevard, Arlington, Virginia 22209

    David Emery, the new deputy director of the Arms Control and Disarmament Agency, has consulted a variety of government seismologists and reached a different conclusion about the best available geophysical evidence. In congressional testimony on 17 May, Emery said that “by far and away the great majoriy of detonations that have occurred have been in a range which leaves little or no doubt that those particular shots have been within compliance. . . . I am convinced that there is no conclusive proof [that] the Soviets have violated” the treaty.


    Nuclear Test Yields

    In replying to a briefing by R. Jeffrey Smith (News and Comment, 17 June, p. 1254) about a recent American Geophysical Union symposium on the verification of nuclear test bans, Ralph Alewine and Thomas Bache (Letters, 29 July, p. 418) make several statements that we, the coorganizers of the symposium, believe are misleading or incorrect. In an invited paper, Alewine and Bache presented the views of the U.S. Department of Defense on two issues: has the U.S.S.R. complied with the 150-kiloton limit set by the as yet unratified Threshold Test Ban Treaty (TTBT) of 1976, and can a comprehensive test ban treaty be verified with high reliability? Their conclusion that many Soviet tests since 1976 exceeded the 150-kiloton limit is at odds with that of many other speakers at the symposium and with the views of many members of a panel that debated that issue at the conclusion of the symposium.

    l’he seismologists at the symposium agreed that attenuation of shortperiod seismic P waves is much less for waves leaving the main Soviet test site in eastern Kazakh than it is for waves from the Nevada Test Site, the source of most U .S. calibration information on seismic magnitude and yield. The disputes were about the size of that bias and about the observed body-wave magnitude (mh) values for specific explosions. Use over the past 15 to 20 years of uncalibrated mb magnitudes for Soviet explosions resulted in overestimates of these explosions by factors of 3 or so. The residual disagreements are minor compared to those of adecade ago, but they still provide the basis for our contrary assertions about probable Soviet behavior.

    Alewine and Bache presented yield estimates at the symposium (which they cite in their letter) derived from preliminary estimates of magnitudes by the U.S. Geological Survey. To our knowledge, they have never used and would never use such uncorrected data in any classified discussion of yield. The USGS magnitudes are rounded to the nearest 0.1 mb unit (equivalent to a 1.25 to 1.5 factor of uncertainty in yield estimate at yields of 150 kilotons) and are not corrected for differing station distributions and thus different station magnitudes.

    We have recalculated magnitudes for all of the larger Soviet explosions in eastern Kazakh from 1976 to 1982 for those effects. Seven explosions since June 1979 have magnitudes very close to and statistically indistinguishable from 6.20.

    After recalculation, none of the magnitudes are as large as 6.3, as in the original USGS estimates. If a value of mh bias of 0.4 is used rather than the 0.3 used by Alewine and Bache and magnitudes are not rounded to the nearest 0.1 unit, the yields of the nine explosions they cite as being of greater yield than 150 kilotons either drop below that value or are very close to it. There certainly are none that approach the value of 315 kilotons given by them at the AGU symposium for one of these explosions.

    Alewine and Bache assert that an mh 2 6.2 in all U.S. experience is associated with yields of 600 to 800 kilotons or greater. While this statement does indeed apply to hard rock explosions at the Nevada Test Site, it is not a correct statement of total U.S. experience. The United States exploded three devices on Amchitka Island in the Aleutians. The observed mb data for these three events, when compared with data from the Nevada Test Site explosions, predicts a yield of 150 kilotons or slightly greater to be associated with an mb of 6.2. Other geophysical data confirm that the Aleutian Islands area is characterized by low attenuation, as suggested by the observed mb-versus-yield relationship.

    Also, the observed mb value for the SALMON explosion in salt in Mississippi indicates that a 150-kiloton explosion in the eastern United States in hard rock would be expected to have an mb value of about 6.2.

    We are convinced that careful calibration using P waves provides estimates of yields of the large post-1976 Soviet explosions (near 150 kilotons) consistent with estimates obtained using other wave types. We believe that the assertion that has been made about alieged Soviet cheating on the TTBT, either in leaks to the press or in published documents by the U.S. government (I), flows from an incorrect calibration of seismic data. When one gets down to seismological details and away from seismological rhetoric, the disagreement in mh bias is 0.1 mb unit, not several tenths.

    Finally, Alewine and Bache state the demonstrable truth that magnitudes of Soviet tests at the Kazakh test site have increased by about 0.3 since negotiation of the TTBT. Given that fact, they imply that Soviet failure to test to the limit for the first years of the TTBT suggests Soviet misbehavior and is thus an added cause for concern. Jack Anderson, in his syndicated column of 10 August 1982, presented the interpretation that the Soviets were testing to the 150-kiloton limit from the start of the TTBT and that their later conduct was based upon penetration of U.S. security and their immediate exploitation of this penetration (knowing the United States had adopted a presumably fallacious mb bias for Kazakh) by raising all yields to our new mb threshold for Kazakh. Why the Soviets would so obviously and for such little gain display the major fact of their penetration of U.S. security is not explained. Also, it is not pointed out that only after this increase in yield did mh values (surface wave magnitudes) of the largest Soviet explosions at Kazakh finally reach those observed at all other test sites in the world for yields of 150 kilotons.

    We suggest a scenario controlling Soviet conduct that seems much more realistic.

    First, Soviet conduct before the TTBT was to make essentially no large weapon tests (above 50 to 75 kilotons) in eastern Kazakh, but rather at Novaya Zemlya. Why did they do this? For the same reason that the United States conducted its multimegaton tests on Amchitka- to prevent excessive ground-shaking in nearby cities (for the United States, in Las Vegas and Reno; for the U.S.S.R., in Semipalatinsk). A fact well known to seismologists but possibly not so well known to intelligence experts is the grossly different attenuation of horizontally traveling short-period waves that occurs in different parts of the continents.

    At a range of about 150 kilometers (the approximate distance of Las Vegas from high-yield tests in Nevada and of Semipalatinsk from the Kazakh test site), the differing attenuations in Nevada and the eastern United States would result in a 75-kiloton explosion in the eastern United States causing about the same level of ground motion as would a megaton explosion in Nevada. Several geophysical criteria indicate that attenuation characteristics in Kazakh are similar to those in the eastern United States.

    Therefore, we suggest that pre-TTBT behavior of the U.S.S.R. was to accept in Semipalatinsk about the same level of explosion-induced ground motion that the United States was willing to accept in Las Vegas.

    Another fact that seems apparent from observed mh values for Soviet explosions at Novaya Zemlya is that few had yields as low as 150 kilotons. One would conclude from the pattern of observed mb’s that this yield was an unimportant reference point for them. So when they entered the TTBT, they then had to make a decision about whether to maintain the Novaya Zemlya test facility for yields of 75 to 150 kilotons only, or to essentially shut it down while accepting somewhat higher ground motions in Semipalatinsk. The magnitudes of the largest Soviet explosions in a given time increased in at least three steps from 1976 to 1979 and have remained constant since. One could interpret this gradual increase in yields at Kazakh as a deliberate and careful evaluation of the acceptability of the higher values of ground motion resulting from higher yield tests.

    In fact, it would have been surprising if the Soviets had immediately started testing at the 150-kiloton limit. We cannot, of course, guarantee that this is the actual rationale explaining Soviet behavior, but it certainly makes more sense than that suggested by Alewine and Bache.

    Soviet violations or possible violations of the TTBT have been cited as one of several examples of the U.S.S.R.’s not living up to various arms control agreements.

    This may be a propitious moment for the U.S. Congress to seek an independent review of scientific issues related to the threshold and comprehensive treaties.

    JACK F. EVERNDENPost Office Box 174, Dav~nportC, alijornia 95017

    LYNN R. SYKES Lumont-Doherty Geologic.t~l Ohservt~toryo$ Columbia University, Palisades. New York 10964 References I . Department of State, Srcurity arid Arrns Coritrol: The Srnrch fbr u Morr Stable Pracr (Government Printing Office, Washington, D.C., 1983).

  2. yale (History)
  3. abcd (History)

    This is hardly surprising. When has Richard Perle – or anybody who hangs around him, for that matter – been familiar with or reliant on evidence?

  4. CKR (History)

    I recall from back a few years a tv program (News Hour?) with several folks sitting around discussing the famous aluminum tubes for Iraq. Two of the folks were Richard Perle and David Albright.

    Albright pointed out that the DOE centrifuge experts had excluded centrifuges as a possibility for the aluminum tubes.

    “Ah, but some experts believe they can be used for centrifuges,” said Perle conspiratorially at a time when we were beginning to understand that the aluminum tubes had nothing to do with Iraq’s nonexistent WMD.

    To Perle, expertise means nothing; propaganda value everything.