George N. Lewis and Theodore A. Postol
On September 17, 2009, the Obama administration announced that it would shelve the Bush administration’s European missile defense system and replace it with an entirely new missile defense architecture. This decision to stop the deployment of 10 interceptors in
Less than five months later, in February, the Obama administration produced an extensive elaboration of the September decision in a document called the Ballistic Missile Defense Review Report. The report asserts that ballistic missile defense technologies have already produced a reliable and robust defense of the
However, a review of the actual state of missile defense technologies reveals that this new vision put forth by the report is nothing more than a fiction and that the policy strategy that follows from these technical myths could well lead to a foreign policy disaster.
With regard to current missile defense technologies, there are no new material facts to support any of the claims in the report that suggest that the United States is now in a position to defend itself from limited ICBM attacks or that any of the fundamental unsolved problems associated with high-altitude ballistic missile defenses have been solved. In fact, as this article will show, the most recent ballistic missile defense flight-test data released by the Department of Defense and the most recent failed test of the ground-based missile defense system in January show quite the opposite.
The Report’s Promises
According to the missile defense report, the continental United States is “now” and for the “foreseeable future” protected against limited ICBM attacks.[1] The report further asserts that this “advantageous position” is the result of well-informed “investments” made over the past decade by the Clinton and Bush administrations in the ground-based midcourse ballistic missile defense (GMD) system, which, according to the report, currently protects the continental United States from ICBM attack.[2]
In the area of regional missile defenses, the report asserts that “recent successes” have demonstrated that the
According to the report, the SM-3 Block IA has been proven highly reliable in numerous flight tests and will be immediately deployed. Under the administration’s schedule, an upgraded variant, the Block IB, will be deployed in 2015. It is to be followed in 2018 by an even more capable Block IIA, and in 2020 by a yet more capable Block IIB.
Because the SM-3 tests have been so successful, these new variants of the SM-3 will be able to accomplish a wide range of major regional ballistic missile defense missions, including enhancing the already in-hand ICBM defenses of the continental
The basic plan for the already functioning GMD ICBM defense will eventually be 30 silo-based interceptors in two existing silo fields—26 at Fort Greely, Alaska, and four at Vandenberg Air Force Base in California. A third field of 14 additional silos will be built as a “hedge” against an unexpected need for additional interceptors.
In addition, the SM-3 and its modernized variants will be widely deployed on ships and on land, in the latter case using ship launch systems that have been modified for land deployment. Elaborate communications and command and control systems will link radars on land and sea with space-based infrared early-warning systems, creating a highly flexible integrated global missile defense with components that can be quickly moved and concentrated as circumstances dictate.
The report, apparently derived from 10 months of intense technical analysis by the Defense Department, therefore lays out a vision of how the
However, the Defense Department’s own test data show that, in combat, the vast majority of “successful” SM-3 experiments would have failed to destroy attacking warheads. The data also show potential adversaries how to defeat both the SM-3 and the GMD systems, which share the same serious flaws that can be readily exploited by adversaries. The long record of tests of the GMD system, and the most recent test in January of this year, shows that it has only been tested in carefully orchestrated scenarios that have been designed to hide fundamental flaws and produce appearances of success. The report provides no material facts or allusions to facts that indicate any technical advances that would counter the long record of orchestrated and dumbed-down missile defense tests.
The proof of these flaws is in the data that the Defense Department cites as evidence of the robustness of the GMD and SM-3 systems. That should be a strong warning to policymakers who believe that the missile defense systems promoted in the report will actually discourage future adversaries from pursuing ballistic missile programs.
The New Architecture
The new plan will depend on a globally distributed system of radars and surveillance and communications systems, which mostly already exist, to provide detection and tracking information needed to guide SM-3 and GMD interceptors to their intercept points.Once interceptors are launched and guided to designated intercept points, interceptor on-board infrared sensors try to find, home in on, and destroy enemy warheads by direct impact. The GMD system uses large interceptors weighing about 50,000 pounds and costing roughly $70 million each. These interceptors will be launched from underground missile silos at
In order to understand how the SM-3 system is supposed to work and how it could fail, it is necessary to understand the many steps that the system must perform when it is in use.
When a ballistic missile is launched,
In some situations, the missile defense system might also have high-flying unmanned airborne vehicles (UAVs) fitted with infrared sensors to track ballistic missiles both during and after burnout. Once the powered flight of a missile ends, however, countermeasures could be instantaneously initiated to prevent these airborne sensors from identifying the warhead.
In the case of launches from
The forward-based X-band radars will have only a modest ability to discern differences in the radar signals from different objects deployed by ballistic missiles at the end of their powered flight. For that reason, these radars will not be able to guarantee that warheads will be confidently distinguished from pieces of debris or decoys. The radars will be able to observe at a range of thousands of kilometers the bodies of rockets that launch warheads, but the radars will have little or no capacity to track warheads deployed by these rockets at these ranges, as the shape and geometry of such warheads make them inherently stealthy relative to the missile bodies.
If ballistic missile trajectories rise above the curved earth into the line of sight of any low-frequency, low-resolution giant
The necessarily small size of the radar antennas on Aegis-equipped ships and the low power of these radars typically result in detection and tracking ranges against warheads and missiles that are too short to allow adequate time for SM-3 interceptors to reach their targets. The new defense architecture attempts to address this problem by assuming that ships will launch their interceptors before their Aegis radars actually observe attacking targets. In many actual engagements, ships would likely never see the inherently stealthy warhead targets with their radars. However, if the external tracking radars have provided the ships with sufficiently precise tracking information, such “blind launches” could be used to guide interceptors to the minuscule volumes of space, roughly 10 kilometers on a side, where interceptors might then be able to use their infrared sensors to find and home in on target warheads.
If an adversary deployed thousands of wires on slightly different trajectories along with warheads, the early-warning radars would not be able to determine which radar signal was from a warhead and which was from a wire. The Aegis ships then would not have the precise tracking information they would need to make a blind launch. This same strategy could also be implemented, with minor adjustments, against the much higher-resolution but inherently shorter-range X-band radars that are also supposed to provide precise tracking data as part of the new architecture and against any airborne infrared sensors carried by UAVs that might, by chance, be in a position to observe the complex of objects launched by missiles.
Thus, any of the many simple countermeasures that disrupt the ability to provide precision tracking data to the Aegis ships could make it impossible for the ships to execute a blind launch. The same kind of basic engagement problems also apply to the GMD system.
Hitting Warhead Targets
In circumstances in which the Aegis ship has sufficiently precise tracking information from external radars for a blind launch, the SM-3 would be launched toward the volume of space where it can then use its on-board infrared sensors to locate and home in on target warheads.
Hitting the warhead once it is “acquired,” i.e., located by the interceptor, is a relatively easy task, but locating the warhead is by far the most demanding task for both the SM-3 and GMD systems. The warhead must be found, identified, and located precisely, and it must be directly hit if it is to be destroyed by impact. Experience shows that hitting parts of a missile’s airframe, even when the warhead is still attached to it, will not destroy the warhead or prevent it from continuing on a nearly unchanged trajectory toward its target.
The three stages of launched interceptors would burn for more than a minute, placing the third stage and its kill vehicle higher than 80 kilometers on a trajectory toward the volume of space where the target is expected. At third-stage burnout, the kill vehicle is released and performs final homing maneuvers for about 30 seconds before it arrives at the selected target, assuming that the system has been able to select the right target or find the location of the warhead on a selected target.
The SM-3 kill vehicle is designed to hit the target at a relatively low closing speed of about four to five kilometers per second and to acquire and home in on targets at ranges of less than 150 kilometers. At this range, the objects in the search volume look like points of light to the infrared sensor on the kill vehicle, so it is not possible for the kill vehicle to obtain information about the shape or size of different objects ahead of it. These substantial limits on what the SM-3 kill vehicle can see makes distinguishing the warhead from other objects a considerable challenge.
The effects of these challenges can be clearly seen in SM-3 intercept test data made public by the Defense Department.[4] In eight or nine of the 10 SM-3 intercept tests from 2002 to 2009 involving these relatively slow closing speeds, the SM-3 kill vehicle failed to hit the warhead target directly. This means that, in real combat, the warhead would have not been destroyed but would have continued toward the target and detonated in eight or nine of the 10 SM-3 experimental tests.[5] Yet, the Missile Defense Agency (MDA) has reported these 10 tests as “successful” without explaining that the test outcomes would not have resulted in true combat intercepts.[6]
The flight-test data, taken from videos published by the MDA, are shown in Figure 1. Each of the images is the last video frame taken by the interceptor just before it hit the target. The flight-test data show that the SM-3 kill vehicles in these tests almost always missed hitting target warheads.[7]
The details of the process by which the kill vehicle tries to identify and hit the warhead make clear why the task of directly hitting the warhead is so difficult and prone to catastrophic failure in real combat conditions.
One to two seconds prior to impact, the images on the SM-3 kill vehicle’s sensor look like slightly elongated dots at the center of the screen. If the kill vehicle hits the body of the rocket, the kill vehicle will tend to shatter and pass through the rocket body much like a bullet hitting a thin-walled drinking glass or an empty soda can, leaving the warhead undamaged and still falling on a nearly unchanged trajectory toward its target.
The flight-test data from the 2002-2009 tests show many striking artificialities that would not be present in actual combat conditions. There are not multiple objects in the threat volume, there are large fins on the back end of the target missiles, the target missiles are always side-on to the interceptor, and the exact geometry of the target missile is known. All these factors considerably simplify the interceptors’ job. Yet, in spite of these artificial advantages built into the tests, the Defense Department’s own data show that the interceptors almost always failed to achieve necessary hits on the warheads.
These test data show potential adversaries such as
Figure 2 shows a very simple countermeasure using rocket technologies that
By using simple explosive techniques to cut the one-stage rocket-target into multiple pieces, a potential adversary could substantially further increase the chances that an SM-3 or GMD interceptor would miss the warhead. Iran and North Korea successfully demonstrated this cutting technique when they separated the stages in the multistage rockets they have already flown.[8] The same could be done to the upper stage of a multistage rocket to counter the homing of the GMD kill vehicle, creating the same confusion of objects to conceal the true location of the warhead from the GMD system.
The scenario illustrated in Figure 2 understates the complexity of the scene that would have to be analyzed by the homing kill vehicle, as the images were generated by assuming that the fragments only tumble in the plane perpendicular to the line of sight of the approaching interceptor. It also does not assume that additional false targets have been created by balloons or unfolded objects that might be deployed as part of this countermeasure.
In the case of the GMD system, which is designed to be able to hit ICBM warheads, the problem is essentially the same. Because the sensor must work at long range, there is little time during the homing process to analyze complexes of multiple targets that could be intentionally and easily created by adversaries. In these situations, the closing speeds will be much higher than those encountered in SM-3 tests, about 12 to 15 kilometers per second compared to four to five kilometers per second. The higher speed requires that the kill vehicle see its targets at much longer range, 450 to 600 kilometers. In order to provide adequate time to maneuver to hit the target, the kill vehicle must have a much larger optical aperture to collect signals from the more distant targets and a much narrower field of view (about 1 degree instead of the roughly 3.5 degrees used in the SM-3 kill vehicle) to be able to get comparably accurate spatial information. In other words, the vulnerabilities of the SM-3 and GMD kill vehicles to countermeasure technologies that have already been demonstrated by
The same fundamental system vulnerability that led to the failure to hit warheads in the SM-3 tests also led to the failure of the X-band radar in the January 31, 2010, GMD missile defense flight test, the FTG-06. The source of this fundamental system vulnerability is the inability of ground-based long-range radars and interceptor-based infrared homing sensors to provide the kind of accurate and detailed images that make it possible to identify the warheads unambigously. Without such true and unambiguous image data, it is fundamentally not possible to recognize the warhead when it is attached to or surrounded by unexpected objects that also individually appear to be different from what was expected.
On April 6, 2010, Aviation Week & Space Technology reported that the sea-based X-band radar being used in the FTG-06 flight test failed to identify the warhead because it encountered “an unfamiliar threat scene,” the set of objects observed by a distant sensor.[9] In the case of an ICBM, it might include a nose cone, a warhead, the upper rocket stage or pieces of the upper rocket stage that were created by an adversary who intentionally cut the stage into pieces, balloons that are spherical, or shaped like warheads, and the like.
In the case of the FTG-06 test, the spent solid-propellant upper rocket stage unexpectedly expelled chunks of rocket materials that created numerous unforeseen radar signals comparable to those expected from the warhead. The radar “scene data” were passed to computers that were programmed to look for a scene that was expected. Because the scene was totally unexpected, the computer analysis failed completely, resulting in a failure to identify the warhead and possibly even a failure to track the entire complex of targets properly.
Because the false radar signals were created by objects that were smaller than the warhead, if the radar had been properly programmed, it could have removed the confusing signals from the small objects before they were passed to the “scene recognition” process. This would not be possible if the objects had been intentionally created by an adversary to have the same length as the warhead or if the warhead had been made to appear different from what the radar expected to see.
According to the Aviation Week article, the GMD kill vehicle observed the target complex with its on-board infrared sensors and picked out the warhead. As will be discussed shortly, however, this fact does not mean the GMD kill vehicle could not have been defeated in the same way as the X-band radar and the SM-3 kill vehicle.
The FTG-06 test illustrates that no matter what sensor is being used, radar or infrared, if the missile defense system knows exactly how the warhead appears to the sensor, then the system can potentially identify a warhead among many other objects. It also illustrates that the appearance of the warhead must be exactly known, and that the warhead must look distinctly different from the other objects.
If the other objects look similar to the warhead or if the warhead looks different from what is expected, the warhead can only be selected as a target by pure chance. Even if the warhead is correctly selected, hitting it may be problematic if it is attached to or enclosed in something that makes it not possible for the kill vehicle to determine where it must arrive to hit the warhead directly. The adversary can easily, perhaps inadvertently, change the scene and target appearance using simple measures, like cutting the upper stage into pieces. The adversary can also change the appearance of the warhead by covering it with radar-absorbing materials, surrounding it with a balloon, or other methods, with totally devastating consequences for the defense.
The failure of the FTG-06 test illustrates the fundamental vulnerability to catastrophic failure of the GMD, SM-3, and all similar such high-altitude defense systems that operate in the near-vacuum of space, about which the authors have been writing for more than a decade.[10] During the first two flight tests, known as the IFT-1A and IFT-2 tests, in June 1997 and January 1998, certain decoys looked enough like the warhead to make it impossible to identify the warhead reliably. In response, the MDA concealed the problem and removed all the decoys that were identified as effective from all subsequent missile flight tests.[11] Now, more than 10 years later, the same fundamental flaw in the GMD system is again revealed, in this case by false targets that were unexpectedly expelled from a solid rocket motor. Notably, the MDA has still not conducted a single GMD intercept flight test against the same combination of warhead and decoys used in the IFT-1A and IFT-2 tests.
Unless the Defense Department can demonstrate convincingly to the world, friends and adversaries alike, that it can deal with such simple countermeasures, no informed adversary or ally will or should believe that either the SM-3 or GMD interceptors will be as robust and reliable in combat as asserted in the missile defense report. The strategy proclaimed by the report rests on assertions that the
If the missile defenses deployed by the
Thus, the Defense Department’s ballistic missile strategy assumes the existence of adversaries sophisticated enough to build nuclear weapons, ballistic missiles, and missile defense countermeasures, but not sophisticated enough to understand that current
The New Emphasis on the SM-3
The domestic implications of the decision to aggressively expand and modernize a flawed SM-3 ballistic missile defense system and to uncritically continue expanding the GMD system are already becoming clear. The Obama missile defense plan creates a framework for putting forward unquestioned and ill-considered rationales for more interceptors and expanded missile defense systems. It will foster an environment of constant lobbying for more interceptors and more sensors to support them. How far this process will go is unknowable at this time, but the indicators of pressure toward uncontrolled and unjustified system growth already exist.
For example, according to the fiscal year 2009 budget, the United States was initially planning to procure a total of 147 SM-3 Block IA and IB interceptors, 133 of which were scheduled for deployment on ships that had been modified to have Aegis missile defense systems. In June 2008, the MDA was already suggesting in congressional briefings that the number of SM-3 interceptors should be increased to 249. In July 2009, the Senate Armed Services Committee upped the ante by raising the number of SM-3 interceptors to be procured in the 2010 defense authorization bill from 147 to 329. Rear Adm. Alan “Brad” Hicks, Aegis/SM-3 program manager for the MDA, laid the groundwork for even higher numbers of interceptors in a January 2009 public meeting by claiming a need for 450 to 500 SM-3 Block IA and IB interceptors. Meanwhile, in August 2009 the U.S. Navy decided that it should upgrade all of its 60-plus DDG-51-class destroyers to have ballistic missile defense capabilities.[13] On April 15, 2010, the current director of the MDA, Lt. Gen. Patrick O’Reilly, testified before the House Armed Services Committee that the Pentagon was planning to procure 436 SM-3 Block IA and IB interceptors for the Navy and 431 THAAD interceptors for the Army by 2015.[14]
The requirement for so many interceptors indicates, at a minimum, three important hidden and questionable assumptions about the SM-3 system. The first assumption is that it is militarily useful to commit one or more interceptors that cost an estimated $10 million each to intercept low-accuracy 1,000- to 2,000-pound conventional bombs that have been launched against unspecified targets or open areas. The second assumption is that these interceptors would actually have a good chance of hitting and destroying the targeted warheads. The third is that the SM-3 could have a meaningful chance of accomplishing the enormously more difficult task of intercepting a mass attack of ballistic missiles. As has already been shown, the last two assumptions wrongly presume that warheads would reliably be destroyed even if the interceptors are able to hit attacking missiles routinely.
Conclusion
If policymakers decide that a strategic defense system should continue to be a central part of the U.S. approach, there are alternative defense systems that could defend the United States from ICBM attack from
By deploying ballistic missile defenses that are easy to defeat, the
The
The negative effects of a costly and energetic
If future arms reduction efforts with
In general, the new missile defense architecture will produce serious doubts about the reliability of small nuclear forces for deterrence. These doubts are unjustified by detailed technical analysis of the true capabilities of these systems, but they will occur and could produce impenetrable new barriers to further nuclear arms reductions.[17] None of these unwanted outcomes need to be a result of the current Obama plan, but without a judicious and careful national assessment of the capabilities and limitations of these ballistic missile defense systems, the pressure to expand them will be both tremendous and without rationale. This new missile defense program could then lead to the usual results: gigantically expensive systems that have little real capability but create uncertainties that cause other states to react in ways that are not in the security interest of the
George N. Lewis has a Ph.D. in experimental physics and is associate director of the Peace Studies Program at
ENDNOTES
1. U.S. Department of Defense, “Ballistic Missile Defense Review Report,” February 2010, pp. iv and 15 (hereinafter BMD Review).
4. Many of the videos from which the image data in figure 1 were derived were initially obtained from the Missile Defense Agency (MDA) Web site. See www.mda.mil/system/aegis_bmd.html. As of the writing of this article, the site has been changed to have one video with data for the FTM-17 test only. For the same video and the data from the other experiments in figure 1, see www.youtube.com.
5. For details on the tests, see http://web.mit.edu/stgs.
6. The MDA published a list of the successful SM-3 “hits,” but the list does not explain that “hit” does not necessarily mean that the warhead would have been destroyed. The preface to the list states, “Since the first intercept test conducted in January 2002, the Missile Defense Agency’s Aegis Ballistic Missile Defense element of the overall Ballistic Missile Defense System has demonstrated 20 hit-to-kill intercepts [emphasis added] out of 24 at sea firing attempts, including two intercepts by two interceptors during one test.” See MDA, “Aegis Ballistic Missile Defense,” n.d., www.mda.mil/system/aegis_bmd.html.
7. This misleading omission in reporting is similar to what happened following the Persian Gulf War. According to the Army’s testimony to Congress, a successful “intercept” meant that a “Patriot and a SCUD passed in the sky.” See “Patriotisms,” Science, April 17, 1992, p. 313. The Army’s initial claim in congressional testimony of a 96 percent intercept rate was later shown by the authors of this article to be “almost certainly zero” as defined by destruction of the SCUD warhead. Even today, the Army claims a Patriot “success rate” of more than 40 percent in
8.
9. Amy Butler, “GBI Test Failure Result of Two Problems,” Aviation Week & Space Technology, April 6, 2010.
10. Theodore A. Postol and George N. Lewis, “We Can’t Tell the Missiles from the Mylar,” The New York Times, July 7, 2000, p. A19.
11. William J. Broad, “Antimissile Testing Is Rigged to Hide a Flaw, Critics Say,” The New York Times, June 9, 2000, p A1.
12. BMD Review, p. 12. For reiterations of this key goal of the administration’s ballistic missile defense strategy, see ibid., pp. vi and 31.
13. Ronald O’Rourke, “Sea-Based Ballistic Missile Defense—Background Issues for Congress,” CRS Report for Congress, RL33745, October 22, 2009.
14. Lt. Gen. Patrick J. O’Reilly, “Fiscal Year 2011 Missile Defense Programs” (unclassified statement before the House Armed Services Subcommittee on Strategic Forces, April 15, 2010).
15. One potential alternative system is to use a small number of stealthy drones that carry very fast interceptors to shoot down cumberome liquid-propellant ICBMs shortly after launch. These are the only long-range missile threats that can be deployed in the foreseeable future or likely ever by
16. Peter Baker, “Twists and Turns on Way to Arms Pact With
17. William J. Perry and George W. Shultz, “How to Build on the Start Treaty,” The New York Times, April 11, 2010, p. WK11.