Back to the Drawing Board: The Need for Sound Science in U.S. Missile Defense

Philip E. Coyle

Clarification made online on February 13, 2013.

The technical core of the U.S. missile defense program is in tatters. Two heavyweight studies in the past 16 months have raised fundamental questions about the science underlying the program.

U.S. missile defense policy has long been controversial. The current U.S. approach has drawn criticism from some who say it is excessive and others who say it is insufficient. The findings of the two studies, however, transcend questions of policy and ideology because they show that the scientific basis for the program is weak. After decades of trying and at least $250 billion spent, the program is in many ways back to square one. This state of affairs has major implications for U.S. national security policy, defense spending, and U.S. relations with Russia, NATO, and allies in the Middle East and Asia.

The more recent of the two studies is “Making Sense of Ballistic Missile Defense,” which was written by a committee under the aegis of the National Academy of Sciences (NAS) and released last September.[1] Thirty-two months in the making, the NAS committee’s 260-page treatise focuses on the intractability of the problem of creating effective boost-phase missile defense—which involves intercepting a missile shortly after launch while its rocket motors are firing—as well as on the difficulties the United States has been having with missile defense in general. In the course of describing the obstacles to boost-phase missile defenses, the report is critical of some elements of the Ground-Based Midcourse Defense (GMD) system currently deployed at Fort Greely in Alaska and at Vandenberg Air Force Base in California and of the European Phased Adaptive Approach being developed for deployment in Europe over the next decade. The purpose of the Alaska- and California-based GMD system ostensibly is to defend the U.S. homeland from intercontinental ballistic missile (ICBM) attack from Iran or North Korea, whereas the purpose of the phased adaptive approach ostensibly is to defend Europe from missile attack by Iran, but the Europe-based system also would include capabilities that would help defend the U.S. homeland from Iranian missile attack.

The NAS panel report, together with an equally important study written by a Defense Science Board (DSB) task force and released in September 2011,[2] leads to the conclusion that both the GMD system and the phased adaptive approach currently lack the basic elements for a viable architecture and, as a result, lack a viable concept of operations.

Both studies also discourage further spending on boost-phase missile defense, notwithstanding nearly 30 years of investment.

Boost-Phase Defense

The NAS committee report is all that decision leaders in Congress and in the Department of Defense should need to stop chasing impracticable boost-phase missile defense projects. The report put it succinctly: The department “should not invest any more money or resources in systems for boost-phase missile defense. Boost-phase missile defense is not practical or cost-effective under real-world conditions for the foreseeable future.”[3]

In agreement with the DSB task force report, the NAS panel’s study explains that all boost-phase intercept systems “suffer from severe reach-versus-time-available constraints.”[4] This means that the defense can run out of time before the enemy missiles are too far away to catch. Although both reports point out that boost-phase technology or its cousin, early-intercept, might be helpful in a crisis under certain circumstances, those situations should not cause policymakers to miss the central point that the time available for boost-phase missile defense can be too short for success.

The NAS panel echoes the conclusions of a 2004 study by the American Physical Society (APS)[5] and the findings of many other experts for years. Nevertheless, the U.S. policy process has kept alive the boost-phase concept and early intercept (intercept before the target reaches apogee).

For example, in a glossy August 2011 handout, the Pentagon’s Missile Defense Agency (MDA) said it would achieve early-intercept capability against medium-range ballistic missiles, intermediate-range ballistic missiles, and ICBMs “from today’s regional threats by 2020 or sonner.”[6] One month later, however, the DSB task force concluded in its report that early-intercept capability is not itself “a useful objective for missile defense in general or for any particular missile defense system.”[7] The task force told the MDA that it was focusing on the wrong thing—early-intercept capability—and losing sight of the basic parameters that determine success or failure in missile defense.

It is remarkable that a Pentagon agency with an annual budget of $10 billion could go so wrong, promising an achievement within a few years that the task force described as “not realistically achievable under the most optimistic set of deployment, sensor capability, and missile technology assumptions.” The Pentagon’s own scientists had to point out how far the MDA had strayed from the basic physics of its systems; the NAS committee has now made the same point.

Midcourse Defense

The first meeting of the NAS panel was held in January 2010; the 15th and last meeting was in July 2011. Yet, because of a review by the MDA, the report was not released for another 14 months, a period almost as long as the time it took the panel to prepare the report.

Last April 30, while the MDA was carrying out its review of the panel report, the committee co-chairmen, L. David Montague and Walter B. Slocombe, provided a shortened version of the report in a letter to Reps. Michael R. Turner (R-Ohio) and Loretta Sanchez (D-Calif.), the chairman and ranking member, respectively, of the House Armed Services Strategic Forces Subcommittee.[8]

At a hearing of the subcommittee on March 6, when she already had learned the conclusions of the report, Sanchez asked Lieutenant General Patrick O’Reilly, the director of the MDA, about the reliability and discrimination capabilities of U.S. missile defenses. O’Reilly pointed to the Precision Tracking Space System (PTSS) as the solution.

Yet the NAS panel’s study, already under review at the MDA, had described the PTSS as “a solution looking for a problem”[9] and recommended to Congress that it be canceled because it “is too far away from the threat to provide useful discrimination data, does not avoid the need for overhead persistent infrared cueing and is very expensive.”[10] Presumably, O’Reilly knew that the study had concluded that the PTSS should be terminated, but he made no mention of that.

Only seven weeks later, in its April 30 letter summarizing its report, the NAS committee was even more blunt, saying that it “finds no valid justification” for pursuing development of the PTSS and “recommends terminating all effort on it.”[11]

Other technical approaches have been proposed for the task of warning the missile defense system interceptors about incoming enemy missiles. One example is the Airborne Infrared (ABIR) system, which is composed of drones, each with a suite of sensors looking up to spot enemy missiles. The DSB task force said the ABIR concept, “although potentially promising as a component that would fit in well with a variety of missile defense architectures and regional situations, is in early development and not ready for inclusion in near-term plans for [phased adaptive approach] architectures.”[12]

The task force pointed out that the technical challenges for the ABIR system “include achieving highly accurate angular accuracy for the sensor as well as packaging (e.g. form factor) to employ on an operational (unmanned) air platform.”[13] This means that the physical orientation of the sensors onboard an unmanned aerial vehicle must be highly precise while tracking the target in flight despite the motion of the platform carrying those sensors. That may explain why the House and Senate appropriations committees zeroed out the ABIR system in their respective versions of fiscal year 2013 funding bills.

Another important missile defense system is the SPY-1 air and missile defense radar carried on all Navy ships equipped with the Aegis system. Four different versions of the SPY-1 are currently deployed on Navy ships. According to the DSB task force report, however, “The current Aegis shipboard radar is inadequate to support the objective needs” of the European Phased Adaptive Approach mission.[14]

Calling attention to the need for better missile defense radars on land and sea, the DSB added that, “[f]or this reason, the TPY-2 land-based radars and the future Navy ship-based Air and Missile Defense Radar…upgrade become critical components of the European defense scenarios.”[15]

The NAS committee supported the DSB task force’s view on the limitations of the SPY-1 radar in a press conference held on September 11, 2012, when Montague explained that the SPY-1 radar was not sufficiently powerful for missile tracking in the European deployment.[16]

Cancellation of Final Phase

The NAS committee recommended canceling the fourth phase of the phased adaptive approach. This phase, scheduled to be deployed in the 2020 time frame, is to be the last and the most technologically advanced, with enhanced capability to defend against potentially longer-range ballistic missiles from Iran, including ICBMs that could reach the United States. This phase, however, is also the most difficult and may not be achievable by 2020 or even much later.

New Expenses for Missile Defense

The recommendations of reports by a National Academy of Sciences committee and a Defense Science Board task force would add large but unknown costs to U.S. missile defense programs. The approaches that the studies recommend would require the consideration of whether new defense spending is sensible on at least the following:

  • A new East Coast missile defense site, perhaps at Fort Drum, New York, or in Maine. As at Fort Greely, Alaska, this new site would require silos from which interceptors could be launched; the supporting infrastructure of command, control, and communications systems; and physical protection.
  • A possible fourth site near Grand Forks, North Dakota.
  • New, smaller two-stage interceptors for the East Coast site that are faster than the existing three-stage, ground-based interceptor (GBI) missiles at Vandenberg Air Force Base in California and Fort Greely.
  • More of those new interceptors to replace the existing GBI missiles at Fort Greely and Vandenberg, and construction of silos in a new missile field at Fort Greely.
  • Development of a new exoatmospheric kill vehicle (EKV)—the part of the interceptor system that would seek out and destroy the incoming missiles—carried on the nose of the interceptor, with far better sensors for tracking and hitting the target. Such an EKV would be more capable and likely heavier than the existing EKV on the existing GBI missiles.
  • The development of a new X-band radar system, roughly twice as big as the existing AN/TPY-2 radar and mounted on a turntable. These modifications would allow it to see further and in more than one direction.
  • Deployment of this new radar at five locations, namely, Fylingdales, United Kingdom; Thule, Greenland; Cape Cod, Massachusetts; Grand Forks, North Dakota; and Clear, Alaska.
  • Permanent deployment of the Sea-Based X-Band Radar at Adak, Alaska, with technical improvements so that the radar can withstand the high winds at Adak.
  • Development of new, more powerful radars to replace the SPY-1 radars on Navy Aegis ships.
  • Development of an Airborne Infrared Surveillance system, not to be confused with the Airborne Infrared (ABIR) system. Although such a surveillance system would be flown on drones like the ABIR system, it would need to provide better targeting information than would be available from the ABIR system as currently defined or from satellites or radar alone.
  • Development of a space satellite system that is some combination of the Space-Based Infrared System and the Space Tracking and Surveillance System to replace the aging Defense Support Program satellites.

The currently defined version of the fourth phase “is not necessary for theater defense and is at best less than optimal for homeland defense,” the committee said in its report. If the first three phases are fully implemented, the additional interceptor capability of the fourth phase “is not required for European (or other theater) defense,” the panel added.[17]

Canceling the fourth phase might help Russia and the United States reach an accommodation over U.S. missile defenses in Europe. The cancellation would eliminate the need for a high-speed interceptor, the Standard Missile-3 Block IIB. Under current plans, that interceptor would be deployed in 2021 in Poland, where it could threaten Russia’s ICBM fleet. According to the NAS committee, such an interceptor would be able to intercept Russian ICBMs launched from bases in southwestern Russia at targets in the eastern United States.[18] For this reason, the fourth phase has been the most contentious part of the phased adaptive approach from the point of view of Russian military and political leaders.

The committee’s recommendation to deploy interceptors at a new East Coast site effectively moves the fourth-phase interceptors out of Europe and replaces them with interceptors on U.S. soil. This would broaden the “battle space,” providing earlier opportunities for intercepts and for multiple attempts at intercepts, but requires the development of a new, faster booster with better target discrimination capabilities than the existing GMD interceptors now deployed in Alaska and California. A new East Coast site would cost billions of dollars more when President Barack Obama and Congress are trying to reduce federal spending, and it presumes that a solution can be found to the age-old problem of distinguishing actual targets from debris, decoys, or countermeasures. “Moreover,” the panel notes, “it finesses the issue of large interceptors close to Russian territory.”[19]

Debris and Decoys

With respect to the need to deal with decoys and countermeasures, the NAS committee wrote that “[t]here is no effective ballistic missile area defense that does not require dealing with midcourse discrimination (or shooting at all potential threat objects!).”[20] “Moreover,” the panel wrote in its April 30 letter to Congress, “early intercept, even if achievable from a forward-based interceptor system, cannot obviate the need for midcourse discrimination, because countermeasures and payload deployment can be achieved very rapidly (as historical experience shows) after threat booster burnout.”[21]

The DSB task force made a similar point about the need to discriminate real targets from debris and decoys: “If the defense should find itself in a situation where it is shooting at missile junk or decoys, the impact on the regional interceptor inventory would be dramatic and devastating.”[22]

Because of the inability to discriminate real targets from debris, decoys, or both and because of the poor record of successful intercepts in tests, especially by the GMD system, the reports from the NAS committee and the DSB task force recommend a “shoot-look-shoot” strategy. This means shooting several times at the same object and looking at it between shots to see if it has been destroyed.

Congressional testimony on this matter by Defense Department officials has been reluctant but clear: the system might have to shoot at each object four or five times to have a reasonable chance of killing that object.[23] If the enemy launches 20, 30, or 50 missiles in a salvo, that could consume hundreds of interceptors.

In battle, those repeated tries would take time. The NAS committee report includes a “Typical Mission Timeline” for a hypothetical four-shot, shoot-look-shoot scenario for intercepting a missile fired from the Middle East at the United States.[24] The first shot is an interceptor launched from Poland cued by an X-band radar in Azerbaijan. Russia has not agreed to a radar at that location, but perhaps Russia and the United States can reach an agreement on missile defense making that possible. That first shot occurs 190 seconds into the flight of the enemy missile.

A second shot from Poland takes place at 526 seconds, cued by a radar at the Fylingdales site in the United Kingdom. A third shot takes place at 772 seconds with an interceptor launched from a notional East Coast site in Caribou, Maine, again cued by the radar at Fylingdales. After each of the first three shots, the radar at Fylingdales performs a kill assessment. At 1,411 seconds, a fourth interceptor is launched from Caribou, once again cued from Fylingdales. At 1,692 seconds, a radar at Cape Cod performs the last kill assessment; and 358 seconds later, the enemy missile hits its target, if the enemy missile has not already been intercepted.

East Coast Site

This scenario demonstrates why the NAS panel recommended a new East Coast missile defense site in northern Maine or perhaps at Fort Drum, New York. According to the report, an East Coast site would provide time for more intercept attempts and for those attempts to occur sooner than would be possible from Fort Greely alone. In theory, successive shots from the East Coast site or Fort Greely would improve the odds.

The NAS committee also suggested a fourth site, perhaps near Grand Forks, North Dakota.[25] This fourth site is reminiscent of the old Safeguard missile defense system, which was developed in the late 1960s and briefly deployed near Grand Forks, with nuclear warheads on 100 interceptors. On October 2, 1975, Congress voted to shut down the Safeguard system after it had been fully operational for only one day because of concerns about the reliability of the system and its cost.[26]

Regardless of whether the new East Coast interceptors would be deployed at two, three, or four sites, the NAS committee’s proposal presumes that technology may be developed to permit the discrimination of real targets from debris, decoys, or both, something the NAS committee and the DSB task force say is essential but not currently in the cards.

The House Armed Services Committee quickly adopted the idea of an East Coast site. In its version of the fiscal year 2013 National Defense Authorization Act, the committee called for the site to be operational no later than December 31, 2015. The full House included this provision when it adopted the defense authorization bill, but the Senate did not.

The Defense Department has been clear that there is no requirement for a third site. In a letter to the congressional defense authorizers, Secretary of Defense Leon Panetta wrote that the House provision “is premature because the administration has not identified a requirement for a third U.S.-based missile defense site, nor assessed the feasibility or cost in a cost-constrained environment.”[27]

The House-Senate conference report on the defense authorization bill called for a study of at least three possible additional missile defense locations in the United States, with at least two of them on the East Coast. The report eliminated the House requirement for deployment by 2015. It also required that the MDA prepare an environmental impact statement for each site and a contingency plan to deploy such a site.

Added Costs

The budget implications of the NAS committee report are enormous and come at a time when defense spending is already under great pressure. Neither the Defense Department nor the Congressional Budget Office has released estimates of how much these additional elements would cost. They certainly would require annual expenditures of billions of dollars; beyond that, one can only extrapolate from the cost of similar systems that have already cost billions (see box, page 10).

Meanwhile, the Obama administration is pursuing two new regional missile defense systems, one in the Middle East and another in Asia. The costs of these systems also have not yet been determined.

The basic architectures of the phased adaptive approach and the GMD system are in doubt because so many of the parts do not work, do not exist, or are not achievable for the foreseeable future. Clearly, a major review and reconsideration is required of all elements of both arrangements. In particular, without a scientifically credible path to effective target discrimination, these projects lack the necessary foundation for a successful missile defense system.

Making sense of all this, sorting through which options might be promising enough to justify further spending and which should be terminated, will be a major challenge for the MDA because the NAS committee and the DSB task force do not agree on what should be done next. The task force study focuses on Europe and on having fast, dependable interceptors; long-range sensors; low-latency (quick-response) communications; and threat discrimination capability. The task force would deploy missile defense assets relatively close to the country possessing the missiles, at which the assets would be aimed as planned under the phased adaptive approach, and does not recommend a new East Coast site, let alone a fourth site near Grand Forks or a new GMD interceptor. It also does not recommend canceling the fourth phase of the phased adaptive approach.

The NAS committee focuses as well on fast, dependable interceptors; long-range sensors; low-latency communications; and threat discrimination capability, but deploys the recommended new interceptors at sites in the continental United States, not in Europe. It limits the scope of the phased adaptive approach.

Neither study comes to a clear conclusion about what should be done with ballistic missile defense sensors in space, and neither one explains how to achieve effective threat discrimination. (Neither study was chartered to conduct such analyses.) A serious study of target discrimination would be a worthwhile project for the NAS and the DSB.

Considering all this, the NAS committee probably should have chosen a different title, as “making sense of missile defense” requires answers to vital questions that were left unanswered, most notably how to achieve effective threat discrimination. The report provides insights, however, into the challenges for missile defense and explains them more clearly than has any document in the public domain since the 2004 APS study.

Expertise Needed at the MDA

The need to constitute a team of scientists with strong research and development capability at the MDA comes through as a top priority in the NAS committee report. The committee explains that the MDA simply could not provide briefings or papers that showed that the agency understood the issues involved and how to make the best scientific and technical choices.

The report said, “Discriminating between actual warheads and lightweight countermeasures has been a contentious issue for [ground-based] midcourse defense for more than 40 years.”[28] From the information that the MDA provided, “the committee learned very little that would help resolve the discrimination issue in the presence of sophisticated countermeasures. In fact, the committee had to seek out people who had put together the experiments” such as the Midcourse Space Experiment and the High-Altitude Observatory 2 “and who had understood and analyzed the data gathered. Their funding was terminated several years ago, ostensibly for budget reasons, and their expertise was lost.” Furthermore, when the committee asked the MDA to provide real signature data from all flight tests, the MDA “did not appear to know where to find them.”[29] Thus, the MDA no longer seems to have the expertise it needs to sort through the many difficult technical challenges that it faces, nor even a record of whom to ask for help.

Summarizing its frustration, the NAS panel concluded that the MDA “has given up trying and has terminated most of the optical signature analysis of flight data taken over the last 40 years. In the committee’s view, this is a serious mistake.”[30]

Without a strong, in-house scientific team, the MDA will not have adequate expertise to decide which path to follow. Private industry will help, but industry also has lost much of the scientific expertise it once had, and the MDA needs to have its own sustained and independent expertise.

As happened with the 2004 APS study, which was too soon forgotten, time has a way of eroding the public policy implications of a scientific study. To scientists, the basic physics is immutable. The public policy arena, however, has a way of discounting scientific findings over time as if to say that the underlying physics is somehow out of date, as if Sir Isaac Newton, watching that apple fall, was wrong about gravity.

In the public policy arena, the fact that someone believes that something such as boost-phase missile defense is viable under real-world conditions, even though for all practical purposes it is not, can be enough to keep it going. Hope and persistence are powerful forces in politics.

If the MDA pursues scientific dead ends and fails to give priority to target discrimination, the United States will continue to have missile defenses that are expensive pipe dreams, cobbled together from components that do not work together in a “system of systems,” failing to satisfy the intended mission.

Philip E. Coyle served from 2010 to 2011 as associate director for national security and international affairs in the White House Office of Science and Technology Policy. From 1994 to 2001, he served as director of operational test and evaluation in the Department of Defense. He worked for 33 years at the Lawrence Livermore National Laboratory on a variety of high-technology programs. He currently is a senior science fellow at the Center for Arms Control and Non-Proliferation.


1. Committee on an Assessment of Concepts and Systems for U.S. Boost-Phase Missile Defense in Comparison to Other Alternatives, National Research Council, “Making Sense of Ballistic Missile Defense,” National Academy of Sciences, 2012, (hereinafter NAS committee report). The version of the report released at a September 11 press conference was in prepublication form. The citations in this article are to the final version.

2. Defense Science Board (DSB), “Task Force Report on Science and Technology Issues of Early Intercept Ballistic Missile Defense Feasibility,” September 2011, (hereinafter DSB task force report).

3. NAS committee report, p. 15.

4. Ibid.

5. David K. Barton et al., “Report of the American Physical Society Study Group on Boost-Phase Systems for National Missile Defense: Scientific and Technical Issues,” Reviews of Modern Physics, Vol. 76, No. 3 (October 4, 2004),

6. Missile Defense Agency, U.S. Department of Defense, “Missile Defense Agency Program Update,” 11-MDA-6310, August 2011, p. 11, Medium-range missiles are generally considered to have a range of between 1,000 and 3,000 kilometers. Intermediate-range missiles have a range of about 3,000 to 5,000 kilometers. Intercontinental-range missiles have a range of more than 5,500 kilometers.

7. DSB task force report, p. 33.

8. Letter to Representatives Michael R. Turner and Loretta Sanchez, House Armed Services Committee, from L. David Montague and Walter B. Slocombe, Committee on an Assessment of Concepts and Systems for U.S. Boost-Phase Missile Defense in Comparison to Other Alternatives, April 30, 2012 (hereinafter NAS committee letter).

9. NAS committee report, p. 119.

10. NAS committee letter. “Overhead persistent cueing” refers to the need for a platform or set of platforms that operate around the clock and provide high-quality information on the launch and trajectory of the enemy target to the interceptor.

11. Ibid.

12. DSB task force report, p. 26.

13. Ibid., pp. 26-27.

14. Ibid., p. 26.

15. Ibid.

16. “The National Research Council Holds a Teleconference on Missile Defense Report,” CQ Transcriptions, September 11, 2012 (copy on file with author). Montague explained:

What the DSB said was the SPY-1 radar is not capable enough to do—support missile intercepts in—in European deployment, which we agree with. SPY-1 is not used for that purpose in the European deployment, a subject that apparently has escaped some people’s read here. The SPY-1 radar is used only for two things. One is to communicate with the interceptor, because the X-band radar is used for, what we call and what the MDA calls, engage on remote. That means all the data tracking data and information that is used to launch an interceptor is—comes from the X-band radar. All the SPY-1 in—in Aegis Ashore does is communicate back and forth with the interceptor.

17. NAS committee letter.

18. NAS committee report, p. 159.

19. Ibid., p. 127.

20. NAS committee letter.

21. Ibid.

22. DSB task force report, p. 27.

23. See Edward Aldridge Jr., Statement before the House Armed Services Committee, March 20, 2003,; Lt. Gen. Henry A. Obering, Statement before the Subcommittee on National Security and Foreign Affairs, House Committee on Oversight and Government Reform, April 30, 2008,

24. NAS committee report, pp. 164-165 (table 5-3).

25. “The committee’s evolved GMD interceptor’s proposed design would use a smaller two-stage interceptor with a total burn time less than a third that of the existing [ground-based interceptor] carrying a larger[,] more capable [kill vehicle]. It would also require adding a third missile field site in the U.S. Northeast and a fourth site in the U.S. North Central states together with additional X-band radars to protect the eastern United States and Canada against Iranian threats.” Ibid., pp. 252-253.

26. For a history of U.S. missile defense systems, including the Safeguard system, see Richard Dean Burns, The Missile Defense Systems of George W. Bush: A Critical Assessment (Santa Barbara, CA: Praeger, 2010), pp. 20-30.

27. Letter to Representative Howard P. “Buck” McKeon, House Armed Services Committee, from Secretary of Defense Leon Panetta, December 11, 2012,

28. NAS committee report, p. 131.

29. Ibid. The 1996 Midcourse Space Experiment (MSX) was a highly successful demonstration of infrared and visible sensor technology in space to identify and track ballistic missiles during the midcourse of their flight trajectory. The High Altitude Observatory 2 (HALO-2) is a modified Gulfstream jet equipped with a variety of ultraviolet, infrared, and visible sensors to observe missile defense flight-intercept tests and collect information about the phenomenology of ballistic missiles in flight. The technology from the MSX and HALO-2 is of interest for improved target discrimination.

30. Ibid.




Clarification: The January/February 2013 article “Back to the Drawing Board: The Need for Sound Science in U.S. Missile Defense” implied that the report on missile defense by a National Academy of Sciences committee warned against the deployment of Standard Missile-3 Block IIB interceptors in Poland because of the capability of those missiles to intercept Russian intercontinental ballistic missiles launched from southwestern Russia. In the passage that the article cited, the committee report was referring to the potential deployment of the proposed Ground-Based Midcourse Defense-Evolved interceptor.