Policing the peace: verifying the Threshold Test Ban.
At the heart of technical discussions about test bans lies the question of verification--how well can each country tell if the other is violating any agreement? Governments continue to develop a host of different technologies -- from satellite surveillance to seismic networks, the United States' main tool for test ban verification -- for keeping an eye and ear on other countries. But in spite of technological and scientific advances in the past four decades and a more recent convergence of views on some scientific issues, the seismic verification question is still hotly debated among a number of scientists and policymakers in the United States. On many issues, "there is an agreement that there is promise," says Willard J. Hannon, a program manager in the Seismic Monitoring Research Program at Lawrence Livermore (Calif.) National Laboratory. "But there are differences as to the extent that the promise has been proven." Some believe that scientists know enough today to adequately verify a long-term treaty, but others disagree.
The first step to verification is detection. With the present state of technology, most experts agree that seismic networks can easily detect events well below the 150 kiloton (KT) limit of the Threshold Test Ban Treaty (TTBT). Indeed, three years ago, seismologists Lynn R. Sykes at Lamont-Doherty Geological Observatory in Palisades, N.Y., and Jack F. Evernden at the U.S. Geological Survey in Menlo Park, Calif., wrote in SCIENTIFIC AMERICAN: "In 1963 [when the Limited Test Ban Treaty was signed] the reliability of measures for the verification of a treaty banning explosions larger than about 1 kiloton may have been arguable, but it no longer is." With its present network of seismic stations, spread over about 35 countries, the United States can detect explosions with yields less than 3 KT (which trigger magnitude 4 or 5 seismic signals) with high confidence, says Robert R. Blandford, a program manager in the Geophysical Sciences Division of the Defense Advanced Research Projects Agency (DARPA) in Arlington, Va.
The big, and politically sticky, issue surrounding the TTBF has been in estimating the yield of a nuclear explosion once it has been detected and identified as such. An incorrect yield estimate, for example, could lead one country to falsely accuse the other of cheating on the TTBT. Or, at the other extreme, it could enable another nation to test weapons that not only violate the TTBT but give that other country a significant military advantage.
The general method that has been used to estimate yield is to derive a relationship between the yields of bombs detonated at the Nevada Test Site (NTS) and the magnitude of certain seismic waves they generate. The problem lies in using this same yield-magnitude relationship to predict the yields of explosions in the Soviet Union and elsewhere. "If we simply applied just this relationship, we'd get enormous yields for the Soviet explosions -- up to 700 and 800 KT," says Thomas Bache at Science Applications International in San Diego.
This is because the underlying geology of the two regions is quite different, and the body waves -- seismic signals that travel down through the mantle and crust -- are affected by this geology. At the NTS region, where the rocks are younger and hotter than those below the Soviet Union, seismic signals are thought to be attenuated more than in the USSR. The result is that a 150 KT explosion might have a magnitude of 5.8 at NTS, but the same explosion at the Soviet test site would measure 6.15, says Sykes, who served on the U.S. delegation to negotiate the TTBT.
Almost every scientist involved now agrees that some sort of correction factor or "bias" must be added to yield estimates. The disagreement is over that number should be used. Based on a classified value for the bias, Reagan administration officials have repeatedly said that Soviet nuclear testing activities for a number of tests constitute a likely violation of the TTBT -- that about 10 explosions each have had yields probably exceeding 150 KT. Sykes has long argued that the government has consistently underestimated the bias. He and many other scientists, both in and out of the government, contend that a more reasonable bias value puts the Soviets substantially in compliance with the treaty. Bache counters that there are not nearly enough data on the Soviet test sites to resolve the question and he points to studies indicating that the U.S. government's value is too large, resulting in estimates that are too low. The Reagan administration still maintains that the Soviet yields probably violated the TTBT, but it seems to be drawing more and more on what it says is "nonseismic" evidence of Soviet cheating.
Administration officials have been careful to use the word "likely" in their statements on Soviet violations because all yield estimates are inexact. Sykes places this uncertainty at 30 percent at one standard deviation, which he says is really not that different from the 41 percent estimate of government sources. Sykes estimates that if the Soviets were to violate the TTBT with a 300 KT test (which is thought to be militarily significant), and given a 41 percent uncertainty in yield, there would be an 84 percent chance that the U.S. seismic network would register the explosion as being at least as large as 212 KT, and there would be a 98 percent chance that the event would be reported as greater than 150 KT. Sykes concludes that these probabilities are sufficiently high to deter the Soviets from cheating.
But the Reagan administration and others apparently disagree and think the accuracy should be improved. The big problem with verification today is that the Soviets have easy acess to geologic and seismic information about the United States while data about the USSR are rarely made available to U.S. scientists, says Bache. "We just make guesses about the USSR," he says. "And then we argue among ourselves." Some of the uncertainty about yield estimates could be reduced if the provisions of the TTBT and PNET, involving the exchange of information on tests and geology, were carried out (see box). But because neither treaty has been ratified by the United States, the Soviets have not been compelled to comply with this aspect of the treaties.
The debate over yield estimates was reportedly rejuvenated within the government this year after two advisory panels to the Department of Defense (DOD) considered the use of surface waves -- seismic signals that travel along the surface of the earth--in addition to the body waves now used in yield estimates of big explosions. In a May 13 letter to DARPA, Thomas H. Jordan, a geophysicist at Massachusetts Institute of Technology and chair of the DOD Technical Review Panel on Nuclear Test Ban Verification, summarized the panel's findings: "Although serious problems have been identified in the procedures currently employed to determine yield from M.sub.s [surface wave magnitude] observations, the Panel is unanimous in its belief that methods for obtaining reliable yield estimates from surface-wave data are now available." The thrust of the panel's report was that the most recent analysis of surface wave data indicates that the body wave bias value presently used is too low, fortifying the view that the official estimates of Soviet yields are too high.
The advantage of surface waves is that, unlike body waves, their propagation is not affected by the properties of the mantle and lower crust. But they can be contaiminated by other surface signals, essentially small earthquakes triggered by an explosion in a process called tectonic release when the crust is in a state of stress. But scientists can tell which signals have been contaminated the least and some maintain that the tectonic signals can than be filtered out.
Bache, however, argues that "no one can be sure that they can correct for tectonic release with any great precision." Others disagree. Jordan thinks that by applying new seismic models and data analysis methods, the accuracy of surface wave measurements can be made comparable to that of body waves. Hannon also thinks that surface waves have a potential for complementing body wave measurements for large explosions but adds that more research is needed to explore how surface waves depend on the geology near the source of the event.
In the midst of all the technical debates and calls for more research, one idea emerges that is openly agreed upon by all involved: There is no such thing as a perfect verification system. There will always be uncertainties in the estimates of yields for a TTBT. And for a Comprehensive Test Ban, there will be always be limits to the minimum magnitude of seismic waves that can be detected and identified as nuclear explosions.
The solution to verifying test bans, as well as to curbing the nuclear arms race -- if there is one--is fundamentally political, say many scientists. How much uncertainty are we willing to live with? How much information are we willing to share in order to minimize those uncertainties? These are questions that ultimately get resolved at the negotiating table, not at the scientists' blackboard.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||part 1; includes related article on the history of test ban treaties|
|Date:||Oct 26, 1985|
|Previous Article:||Patents for seeds and plants.|
|Next Article:||Cell transplants into monkey and human brains.|
|Policing the peace: verifying a comprehensive test ban.|
|Soviets visit U.S. for mock nuclear blasts.|