How Eminent Physicists Have Lent Their Names to a Politicized Report on Strategic Defense
“Physicists Express ‘Star Wars’ Doubt; Long Delays Seen.” This headline in the New York Times of April 25, 1987 heralded a 424-page Report to the American Physical Society of the Study Group on Science and Technology of Directed Energy Weapons. Over the following week the major media uncritically spread the message: the nation’s senior physicists—the ones who have produced the marvels of modernity, the ones to whom we pay hefty tuitions for our children, the very ones who deal in symbols too complex for mere mortals—had risen above the partisan clamor over the Strategic Defense Initiative (SDI), had conducted an impartial study, and had concluded that all the talk of a serious defense against ballistic missiles was nonsense. The physicists’ knowledge of the laws of nature and of the details of technology had dispassionately shown that well-nigh insuperable hurdles stand in the way of building devices that would meet present requirements for antimissile defense. Moreover, even if these requirements were ever met, the Soviets could easily make a few changes in their offensive forces and leave us worse off than before we started.
In fact, however, the American Physical Society (APS) report was not written by an impartial jury of qualified scientists. Although some of the physicists involved with the report have very distinguished records of scientific achievement, not a single one of them (not even Charles Townes of Berkeley, who shares credit for inventing the laser) has ever worked in the practical field of developing directed-energy weapons. Perhaps this helps explain why the report contains so many errors and internal contradictions (telling us in one place, for instance, that the engines of long-range missiles burn for between three and six minutes, while in another place assuming that they burn between two and three minutes). More importantly, the fact that the report was commissioned, written, and reviewed by people who, like Sidney Drell, Wolfgang Panowsky, and Herbert York, have spent much of their time during the past generation advocating the weaponry of Mutual Assured Destruction (MAD), and opposing the alternative represented by SDI, may also help to explain why it is a work of ideology masquerading as science—a political tool the purpose of which is to convince Americans not to try to acquire defensive weapons.
The report consists of a network of assumptions overlaid upon scientific filler materials. Its conclusions are predetermined by the interplay of the assumptions. Because these assumptions are either erroneous or arbitrarily tendentious, or both, the report as a whole offers a particularly striking illustration of the principle of garbage in, garbage out.
One major example is the notion that all warheads are created equal. Thus the report begins with this evaluation of the U.S.-Soviet balance in offensive missilery: “. . . the qualitative status of U.S. and Soviet ICBMs is similar. The Soviets are thought to have a slight lead in silo hardness, the U.S. is ahead in solid propellant technology and in accuracy.” This makes as much sense as comparing rifles and shotguns by talking about the gunpowder they use. In fact, the Soviets have an overwhelming edge in the combination of warhead yield and warhead accuracy. That is why Soviet warheads can destroy America’s vital military targets while American warheads cannot destroy the Soviet. Union’s vital military targets. The authors of the report, projecting upon the Soviets their own commitment to the MAD doctrine, refuse to understand that Soviet ICBMs are not meant to hurl indiscriminate destruction at U.S. society at all costs, but rather are designed to perform a precise military task.
This willful blindness permits the authors to take as given that the Soviet Union, in order to counter U.S. missile defenses, would do things like divide missile payloads into great numbers of nuclear bomblets, and cause missile boosters to burn out within the atmosphere. But while such measures might conceivably increase the chance of getting some warheads through an American defensive system, they could only do so at a severe cost in yield, accuracy, and reliability with regard to any given target. They would thereby make it impossible for Soviet ICBMs to do the job for which they were built in the first place.
So pervasive is the tendency in the report to substitute its own assumptions for reality that when the words “if,” “nominal,” or “assuming” appear, one has to hold on to one’s intellectual wallet. For instance:
For a nominal hardness6 of 10kj/cm2, at a range of 1000km, the brightness-time product needed is 1020j/sr. Assuming that the engagement time is .01 seconds (as noted in calculations that follow later in this chapter) a brightness of 1021w/sr is required . . . If the mission requires a brightness of 1021kw/sr, a laser at [a wavelength of one micron] operating with 10m mirror will need to generate about 30 megawatts . . . at [three microns] 300 megawatts would be required.
English translation: If we assume, without argument, that the Soviet missiles to be killed by U.S. laser weapons are ten times “harder” (i.e., more resistant to lasers) than the hardest missile flying today and if we further assume, again arbitrarily, that we have only one-tenth of a second to make the kill, it then follows that in order to kill missiles the U.S. would have to put into space either a 30-megawatt laser operating at a wavelength of one micron, or a 300-megawatt laser operating at a wavelength of 3 microns. The U.S. does not now have the capacity to produce such lasers, and will not for the near future.
But if we were to say (as we could with much greater plausibility) that a full second is an acceptable time to allow for shooting down the real missiles in existence today, then one would have to conclude that a 10-mega watt space-based laser operating at a wavelength of 3 microns would do a good job of defending against missiles. The authors of the report know that such missile-killing lasers are very much within reach. But since they do not want the U.S. to develop a system of strategic defense, they weave a web of assumptions leading to the conclusion that the lasers we can build are not worth building, and that the ones worth building are not buildable.
In sum, the report’s modus operandi is as follows: Show the near-impossibility of our performing technical feats that the report implies are essential to ballistic-missile defense, but that in fact are irrelevant or imaginary. Then use this to argue that more research must be done into how to meet an even stricter requirement. On the other hand, take it for granted that, on the Soviet side, the most astonishing technical feats—“ballistic” missiles that would cease to be ballistic and jig randomly in their trajectories, and warheads equipped with all-azimuth sensors—can easily be performed to evade any U.S. defensive system.
This modus operandi is neither science nor technology.
A good illustration of how the report deploys the second aspect of this technique is its treatment of the crucial question of missile vulnerability. Obviously the less vulnerable (or “harder”) Soviet missiles are, the more effective must be the directed-energy weapons we design to counter them; and the more effective the weapons must be, the more difficult it will be to build them. Not surprisingly, therefore, the report devotes a good deal of time to creating the impression that the bodies of missiles are very tough, and can easily be made even more resistant. The truth, however, is that real missiles happen to be pretty flimsy things.
Indeed, at least since 1978, it has been known that the amount of on-target power needed by a laser to destroy missiles is not very great. Thus, whereas the materials out of which our own Minuteman and MX missiles are made will burn through after having received about 6,000 watts per square centimeter per second, tests have shown that when those missiles are under the heat and pressure of boost phase, they will stand less than one-sixth of that amount before they disintegrate. That ratio of one to six also holds for missiles made of aluminum—as are the overwhelming majority of Soviet ICBMs—except that the amount of on-target power needed to destroy aluminum missiles has proved to be much smaller (only between 80 and 160 watts per second per square centimeter).
One will find nothing of this in the APS report. The report gives no facts about how hard missiles are. Instead it gives the impression that they are very hard, and discusses various methods for making them even harder: coating missiles with laser-resistant materials; spinning the missiles to spread out the effects of laser beams; and making the missile engines work so fast that they burn out before leaving the atmosphere’s protection.
Coating missiles is often touted as the simplest laser counter-measure. The problems, however, only begin with the weight of the coating. Lowell Wood of the Livermore Laboratory and Gregory Canavan of Los Alamos have pointed out that the report’s distinguished authors made an elementary mistake by assuming that the weight of the coating to be placed on each stage of a missile would be proportionate to the mass of that stage rather than to its surface area. This leads the report to underestimate the weight of the coating, and therefore the number of warheads that the Soviets would have to take off their missiles to “pay” for the coating’s weight. Thus, say Wood and Canavan:
. . . if the Soviets put 6 metric tons of anti-laser shielding on an SS-18—which would represent about a half inch of shielding over the entire skin of the missile—the APS report states that the SS-18 could still carry 5 warheads out of its normal complement of 10. . . . The correct answer, however, is that it would have to take off a total payload mass equal to that of all of the 10 warheads.
To be sure, as Wood and Canavan concede, in removing warheads the Soviets would also remove fuel from the post-boost vehicle. This might make room for one or two warheads aboard the hardened SS-18. Yet even if it could be thus retrofitted, the SS-18, the key to any Soviet first strike, write Wood and Canavan, “would be reduced to something far less . . . by the mere existence of an American strategic defensive capability. . . . Any attrition of the attack by the operation of the defensive system would be an additional benefit.”
But what about the prior question: could the shielding be applied? It just so happens that the materials that are best at resisting lasers—vitreous carbon compounds—are too brittle to take the stresses of a boosting missile. This fact receives veiled acknowledgment—but no more—in the report. Yet even if a good material were found, and the sheets were glued or bolted on, does anyone think that the skin of a missile that is designed barely to hold itself together could support tons of material hanging from its sides? This elementary question is one the report does not even mention, let alone address.
The report is similarly careless on the second counter-laser method—spinning missiles. Never mind that to spin a missile the engines would have to be modified and that the inertial guidance system would have to be redesigned to take into account the new forces acting upon it. But how would the missile’s skin react to the centrifugal force—especially with heavy shielding hanging on? Given all of the forces already working upon the missile, would this new force push it over the edge? To what extent would the weakening caused by centrifugal force negate whatever spreading of the laser beams that the spinning would produce? The report says: “Rotation of missiles at angular rates of the order of 1 rps have been studied and shown to extract little or no penalty to the offense. It is likely that missile rotation can be accomplished on a retrofit basis.” But no one has ever tried to retrofit a missile to rotate. Perhaps that is why the report gives no details of any such “study.”
The third counter-laser method dealt with by the report—the mythical fast-burn booster—has been a hardy perennial of the anti-strategic defense crowd since 1982. That is because such boosters would have three advantages. First, a missile whose engines are not burning cannot be easily seen. Second, it is not under as much strain and hence is not so “soft” or vulnerable. And third, during the time when it is “softest,” the atmosphere protects the missile to some extent from many lasers, and entirely from particle beams.
The report refers to studies by Lockheed and General Dynamics to the effect that no physical principle prevents the design of a rocket that reached sufficient speed in 60 to 100 seconds to send its payload to intercontinental range. Yet neither the U.S. nor, so far as we know, the Soviet Union has tried to build such rockets. The reason is that when a missile’s engines cease burning before leaving the atmosphere behind, the missile cannot impart to its payload the precise thrust and direction it needs for accurate delivery—that is, to be a militarily rational weapon.
The report’s authors were aware of this, but nonetheless wrote a strong endorsement of the fast-burn booster. The only explanation for the contradiction, other than professional irresponsibility, is that the authors discount military utility and that they value highly any factor, however small, that would favor the delivery of any kind of nuclear device at all.
Just as the report strives to create the impression that missiles can become vastly less vulnerable than they now are, or are likely to become, so it strives to convince us that the defensive laser technology we already have at hand is vastly less effective than it actually is.
The report’s discussion of this subject begins with chemical lasers—the kind that everyone acknowledges are closest to being made into weapons. The “bottom line,” stated in the report’s Executive Summary, is that before this can happen chemical lasers must improve by a factor of 100. The text, however, does not argue this point directly. As we have seen, the issue is largely predetermined by the arbitrary initial assumption that missile boosters have a “nominal” hardness of 10,000 watts per square centimeter per second, and that kill times must be one-tenth of one second.
Most of the report’s discussion of chemical lasers is textbook physical chemistry. It says nothing about how close we are or are not to a device of a given power. The last page-and-a-half deals with the engineering of chemical lasers, and is full of hints—but only hints—of unsurmounted or perhaps insurmountable practical problems. These hints add up to the suggestion that it would prove difficult to build a chemical laser much more powerful than the largest one we now have (2.2 megawatts). Yet in 1980-82 American aerospace companies were proposing to build 5- and 10-megawatt lasers. Moreover, the existence of a 5-megawatt laser was reported in Soviet publications as early as 1975! Rather than openly challenging the feasibility of a space-based chemical laser, then, the report raises vague and indistinct doubts.
Of course the question before us is whether or not the U.S. should actually build a 10-megawatt chemical laser and carry out the only really meaningful tests by shooting down missiles with it. This is a political question, not a technical one. And this is the decision that the report tries to influence by insinuating that the technical uncertainties are too great. Yet the truth is that technical uncertainty has not been a factor in this program for many years. Indeed, the chemical-laser program has been intentionally held back. One would never know from the report that the U.S. Defense Advanced Research Projects agency, in 1980, gave to the Carter administration and to the Congress the option of a program to test a prototype 5- or 10-megawatt chemical-laser weapon in space by the end of 1986.
The report also deals with other kinds of lasers that might be used to defend against missiles. Since no one contends that these other kinds of lasers are ready to be made into weapons, the report’s task is merely to make them appear even farther away from feasibility than they are. But even here the report arbitrarily overstates, and once more by a huge margin, the requirements for effectiveness. For example, the Executive Summary says that one type of laser—the Excimer—must operate with a power of a billion watts in order to work as an effective weapon. Yet, figuring backward from the authors’ own estimation of the power required to rupture a missile’s skin, we discover that only some six million watts would be needed.
Similarly with another laser on which we are working—the free-electron type (F.E.L.). For the F.E.L., the report sets a power requirement wildly greater than its own criteria would actually call for. After reciting all of the real difficulties in the program—achieving high power with any degree of stability, getting mirrors to bear up under high power, short wavelength beams—the authors, not content with concluding that F.E.L.’s are not yet ready to become weapons, seem unable to resist the urge to drive a stake into the program’s heart. So they write: “We estimate that for strategic defense applications, a ground-based free electron laser should produce an average power level of at least 1 Gw [billion watts] at 1 micron wavelength—.” The report further assumes a 75-percent loss of power for ground-based lasers (due to relay mirror losses, as well as atmospheric scattering and absorption). But if the reader applies the figure from laser on-target power implied by these assumptions to the report’s own “nominal hardness” for missiles, he cannot but conclude that such a 1-billion watt F.E.L. ground-based laser would still be between 25 and 100 times more powerful than necessary. Is there any reasonable explanation for such a high requirement? None.
Among the more pathetic of the report’s sections is the one dealing with the task of recognizing and tracking missile boosters. The main message in this chapter is that it is well-nigh impossible to track missiles from space with sufficient accuracy, and that it is impossible to discriminate between the warheads and the decoys that these missiles would release if they were not destroyed.
The first part of this message is based on the postulate that in order for a space-based laser (or a mirror relaying a ground-based laser) to hit a missile with a diameter of 1 meter, the detection and tracking instruments, located in orbit 22,300 miles above the earth, must be able to locate with an accuracy of 1 meter or better “the absolute position in space” of every missile.
If one accepts that innocent-sounding premise, one will conclude along with the report that the job would require 40 meter-wide infra-red telescopes 22,300 miles above the earth, and one will gasp at the thought. To get around this enormity, the report proposes an even greater one: fully 6,000 dish-type microwave radars in low earth orbit.
But this is nonsense from beginning to end. If one were to give the job of precise tracking of missiles to the missile-killing laser stations themselves (operating perhaps 500 miles high) it would be wholly unnecessary for them to determine the missiles’ absolute location in space. They would only have to determine the missiles’ location relative to themselves.
Interestingly enough, the original space-based laser program took this common-sense approach back in 1979, and went on to develop the technology to do precisely this in a program called Talon Gold. With this approach the job of detecting and tracking devices from high altitudes would simply be to give each low-altitude station information sufficient for it to point its acquisition sensors within roughly ten degrees of its target. That kind of high-altitude tracking is very much within current technology. The report’s approach amounts to looking for a very hard path when an easy one is evident, and then complaining about how hard the path is. Is this science?
The report’s treatment of detecting and tracking post-boost vehicles (PBVs) is of the same kind. “Even if detection is feasible, imaging as a means to distinguish reentry vehicle (RV) deployments from decoy deployments seems unlikely.” So indeed it is. But distinguishing between warheads and decoys has nothing to do with tracking and destroying the post-boost vehicle. The report goes on to suggest that although American heat-sensing devices could detect the hot exhaust of the post-boost vehicles’ engines, “these emissions could be reduced by using cold gas thrusters, or shrouds.” But this is purely gratuitous. It is conceivable to use cold compressed gas to power PBVs. But because the energy content of cold gas is so low, a PBV that ran on it would have to pack so much gas that there would be no room for anything else.
As for “shrouds”—mythical skirts that would sprout from a rocket to hide its exhaust—the report gives no hint of how they could be built, nor of how the shrouds themselves could avoid getting hot enough to be detected. When something produces heat, that heat has to go somewhere; one would think that eminent physicists would remember that.
The second part of the chapter’s main message is that decoys and other penetration aids are so good that they cannot be distinguished by observation, no matter how accurate. One looks in vain through this part of the chapter for any evidence to support its conclusions. Instead, the reader learns that the conclusions are based on “general considerations.” Certainly they are “general,” and some of them—e.g., that decoys are perfect by definition—are unbelievably ignorant as well.
Of course, it is easier to define decoys as perfect than to make perfect decoys. How does one fix things so that the temperature and the motion of a body weighing a couple of pounds will match those of a similarly shaped body that weighs one hundred times as much? The report gives no evidence. It just assumes. For example, the report stipulates that the decoys would be spun, just as warheads are spun. But if the PBV imparts the same spinning motion to a tinfoil balloon and to a warhead, these very different objects will necessarily spin with an easily recognizable difference. As for another quality of decoys that supposedly makes them indistinguishable from warheads, their infra-red emissions, or “signatures,” the U.S. has long since found that decoys and warheads give off recognizably different ones.
Finally, there is the question of the survivability of defensive assets. The authors of the report obviously believe that such assets could not possibly survive in peacetime—much less in wartime. Just as obviously, their conviction does not proceed from any scientific or technical facts. It is interesting to see how these eminent physicists try to present their political prejudice regarding anti-missile defense in the guise of scientific judgment.
In peacetime, asserts the report, space-based assets might be subject to “covert attacks.” The authors do not say what would constitute the attack, or how it could be kept covert, or why the space station being attacked would not have its sensors looking for an attack and its counter-measures ready to move out of the way or to repel it. As the authors imagine it, the enemy’s “space mines” would align themselves next to the defensive satellites, and shadow them, ready to blow up—and the defensive satellites would let them do it.
Why? The authors claim that to program battle stations to destroy any satellites that tried such shadowing maneuvers, and actually to enforce a “keep-out zone” around our defensive battle stations, would be “problematic.” Indeed, it was problematic for the U.S. frigate Stark to defend itself against missile attack in the Persian Gulf—but only because U.S. policy had apparently led the ship’s captain not to turn the ship’s defense on “automatic.” There is, then, a problem with keep-out zones. But it is in no way a technical problem.
Later, the authors state that the “space mines” could be “stealthed” so that the defensive stations would not see them coming. But they are adamant that the defensive stations themselves could never be “stealthed”—and they cite the difference in size as one of the main reasons why. Anyone who so writes has clearly never been involved in the “stealthing” of anything. There is no doubt that reduction and alteration of radar and infra-red “signatures” would play an important role in any effort to attack the defense, and in any effort to defend the defense. But the signature modification of aircraft amounts to a contest of detail; the winners are those who actually manage to hide better, and whose sensors prove better at seeing through the other side’s attempt at hiding.
In wartime, the principal threat to defensive stations, according to the report, would be nuclear-tipped missiles sent up directly to destroy them. Through the magic of arithmetic, the authors conjure the following: “If one credits the offense with a nuclear [anti-satellite] attack capable of flying to 1,000 km altitude in 300 seconds, launching 10 offensive interceptors against each platform under attack and deploying 1,000 decoys and 2 hardened warheads from each interceptor, the defense . . . needs to address approximately 30 targets per second and a total of approximately 10,000 targets per battle station.” The point of this is that the laser stations could not begin to defend themselves against so many threats. But why should a laser station built to destroy boosters in a second or so spend its time swatting thousands of decoys nearby when it can destroy the few boosters that carry them?
Based on its arithmetic, the report concludes that “Direct ascent nuclear attacks are so threatening that no directed energy battle stations in low earth orbit can be presumed to be survivable on their own.” But the fact is that regardless of what is packed atop any anti-satellite booster, that booster constitutes one and only one problem for the defensive station. This is not to deny that any one station can be overwhelmed, but even if the offense knew the precise location of each defensive station, a very large number of anti-satellite weapons would be required to overwhelm each station.
The mistakes and the sleight of hand in the report are so many and so important that it is impossible to regard the whole as an honest product. Rather, the report must be regarded as an act of low politics. It is not an objective evaluation of the state of the art of directed-energy weapons. It is a document intended to bolster campaigns to keep SDI in its research-forever deploy-never phase. Yet even as such, it is not a serious argument against SDI, aimed at scientists in the field. Clearly, anyone acquainted with the field—and indeed almost anyone who gave the report a close reading—would see that it does not argue, but rather plants deceptive axioms. Hence we must conclude that the report was not meant to stand up to close reading. Its intention was to dazzle casual readers—mostly the press—and to generate the kinds of headlines that in fact it generated. In other words, the authors took their prestigious academic affiliations and lent them to the service of a political cause.
If this kind of thing had been done—as indeed it has been done many times before—under the auspices of the Union of Concerned Scientists, or the Soviet Committee for Peace against the Nuclear Threat, there would be nothing surprising about it. But this report bears the imprimatur of the American Physical Society, which is supposed to represent the views of America’s physicists; this makes it both shocking in itself and a sign of even worse things. As Frederick Seitz of the Rockefeller University, himself a former president of the American Physical Society, has noted, the report’s willingness to neglect professional standards for the sake of political objectives has ominous precedents:
Physicists with long memories will recall that when the Nazis came into power in Germany in the 1930’s, the German physics journals—which had been, until then, among the finest in the world, as the journals of the American Physical Society are today—began to publish work of questionable quality. That was one of the earliest indications of the decline of German science in the pre-World War II period. One thinks also of the unfortunate depths to which some Soviet scientists, especially in genetics, have descended at various times to satisfy the prevailing currents of political thought in the Soviet Union.
Seitz, however, has been one of the very few physicists who have challenged the report. In fairness one must note that few have read it, and fewer still are acquainted with the field. Moreover, those who are in the field are mostly under contract to the U.S. government. Hence the public had a right to expect the government to issue a responsible refutation of the report. But the government—and specifically the Pentagon’s Strategic Defense Initiative Organization (SDIO)—has limited itself to a lame two-page comment that does not contradict any of the report’s specific points. Its strongest statement is that “we would not have made several of the assumptions they made in defining the technical requirements.”
But why did the Pentagon not lay out what the correct assumptions are, and explain why they are indeed correct? The answer lies in a fault deep in the Reagan administration’s approach to SDI—the very fault that made the report’s thrust possible in the first place. The administration has gone out of its way to define SDI as research into a question it officially regards as entirely open: whether anti-missile defense is possible. Thus, the administration has framed the SDI program as abstractly as possible; it has not come up with a realistic definition of the threat we face, and of the job to be done to meet the threat, which it is willing to defend in public. Without such a definition—without, in other words, an anchor in reality—one set of assumptions is as good as another.
Perhaps the principal reason for this perennial state of suspended animation is that the Reagan administration has sought to defer grappling with the considerable diplomatic, bureaucratic, and political consequences of committing itself actually to building anti-missile defenses. Thus, the administration has not said: Recognizing that time is the key to most human contests, let us build the best defensive weapons we can with what we have today, against the threat that our intelligence can discover, or that we ourselves are incorporating into our own offensive forces. Instead the administration has said: Let us give free rein to our imagination of what the most extreme threat might be in the distant future, and use that as a benchmark for judging the adequacy of our present defensive technology. This is a process without logical end.
The APS report’s capacity for mischief stems from the fact that the basic premises of the SDI program, which it so artfully caricatures, are seen by many to have been dictated by the “best and the brightest” of American science. To this there is a straightforward answer. The President ought to convene a body of experts, just as eminent as the ones responsible for the APS report—but this time people who want the U.S. to have anti-missile defenses—and ask them if any other premises and approaches are scientifically tenable. But, alas, one fears that the Reagan administration would be almost as uncomfortable with such a report as the American Physical Society itself.