|
|
|
|
|
|
|
|
|
|
The Honorable Philip E. Coyle
Senior Advisor
Before the Missile Defense Testing
June 11, 2002
For Official Use Only
Mr. Chairman, and Members of the Committee, thank you for the opportunity to discuss the adequacy of testing of the Ground-based midcourse element of the planned Missile Defense System. I last appeared before this Committee on September 8, 2000, when I was serving as the Director, Operational Test and Evaluation in the Office of the Secretary of Defense. You requested that my testimony today focus on the recommendations I made while in office to improve the adequacy of missile defense testing, the most important of those recommendations and which of those recommendations have been adopted, the risks of not adopting those recommendations, and the likely schedules and capabilities in missile defense that can be achieved in the relatively near term. To have a comprehensive discussion of missile defense has become much more complicated, over the past year and a half, because of the Administration's plan to develop a layered system consisting of boost phase, midcourse, and terminal phase elements with land, sea, air and space-based components. Today there are a dozen or more major defense acquisition systems all dedicated to missile defense, and the Missile Defense Agency (MDA) has not identified an architecture that would tie all these elements together. Accordingly when discussing the Ground-based midcourse (GMD) system, we are looking at only a part of the overall idea.
The recommendations I made two years ago and still make today fall in three important categories:
I'd like to elaborate on each of these.
GMD testing will need to become more realistic to prepare for the IOT&E phase of development. This means tests such as at different times of the day and night, with longer range intercepts and with differing target trajectories, with different types of decoys and countermeasures, with tumbling target reentry vehicles, with multiple simultaneous engagements, with unfavorable meteorological conditions, and with incomplete or inaccurate prior information. For example, while we can debate whether a country, say, North Korea would be reckless enough to launch an ICBM with its tell-tale return address at the United States, if they could do that they would probably launch more than just one. This is why tests with multiple targets will be important.
The GMD program has been attempting to develop such tools and they would tell you they have made good progress. However, this has been an especially difficult area for missile defense. The simulations are technically very challenging to build, have fallen significantly behind schedule, and so far are not particularly realistic. For National Missile Defense to work, the GMD program will need to be able to trust these simulations and rely upon them for engineering design, for mission planning and for training. This is because it will never be possible to demonstrate in open air tests all of the possible combinations and permutations of targets and decoys, trajectories, and environmental conditions that the GMD system might encounter in realistic operational situations. Having good ground testing facilities and simulation tools will be especially important if the Administration's concept of a layered system is ever to be realized. In a layered system there would be a constant interplay between the various elements of the layers. Those interactions would be so complex they likely could never be fully reproduced in open air tests. Realistic simulations will be essential.
One of my recommendations, made now nearly two years ago, was that the pace of missile defense testing would have to be accelerated if the MDA wanted to have any hope of achieving some kind of capability in the next decade. This is because there are roughly 20 developmental flight intercept tests that must be done before the NMD system can be ready for realistic operational testing. If those 20 tests each took about a year, as some of the first flight intercepts have, it could take 20 years before the system would be ready for realistic operational testing. And that assumes all the tests are successful. Then realistic operational testing could take several more years, so you could be looking at a quarter of a century before an effective system would be demonstrated as ready for deployment. To its credit, the MDA has accelerated the pace of GMD flight intercept tests. If the next flight intercept test, IFT-9, is conducted in late July as scheduled, the MDA will have conducted 5 tests in a little over twelve months. However, except for the most recent flight intercept test, IFT-8, fired on March 15, 2002, all of those tests were the simplest of any of the tests planned, and were essentially identical. Where IFT-6 and 7 were to have been more difficult than IFT-3, 4 and 5, they turned out to be about the same. Accordingly the actual progress in the flight intercept tests so far has suffered. Thus, while the number of tests has gone up in the past year, the GMD program is actually behind where it hoped to be at this point by about a year, measured from about two years ago. The challenge for the MDA will be to sustain the current pace of testing while adding complexity and realism. This won't be easy. It also will require adequate spares and backup test systems so that failure or delay in just one area does not delay the entire testing program. The Missile Defense Agency has said that they are trying to do all of the things described in the three categories above, and I believe they are. However, it is not conceivable that they can do all these things and still meet the date of 2004 that Lieutenant General Kadish has testified they can meet for early deployment. 2004 is only two years away! If they start cutting corners and don't follow these recommendations three things will happen. First, the tests - such as they may be - will give members of Congress and the American Public a false sense of security that missile defense will protect them from enemy missiles. Second, if and when the missile defense programs reach realistic operational testing, they will fail for not having been tested before under realistic operational conditions and environments. Finally, the GMD missile defense program will fall hopelessly behind schedule, and the Congress and the American Public will lose patience with the entire development effort and cancel its funding. This has happened before with other military development programs. Mr. Chairman, in your letter of invitation for this Special Briefing, you asked whether the Missile Defense Agency had taken actions to make the testing program less challenging. I don't believe they have. Missile defense is very difficult, and the GMD program is still early in its development. However, the Congress and the American Public could already be misled. For example, the decision of the MDA to restrict and classify information about future flight intercept tests creates a situation where the actual progress may not be what it seems to be from terse reports. Since 1997, the GMD program has conducted eight flight tests. A table is attached to this testimony which summarizes these tests. The first two, named IFT-1A and IFT2, were fly-by tests designed to collect target phenomenology information and not intended as intercept tests. The next six tests, IFT-3 through IFT-8, were all flight intercept tests. IFT-4 and IFT-5 conducted in January and July of 2000 respectively both failed to achieve an intercept which became a principle reason why on September 1, 2000, President Clinton decided not to begin deployment of GMD components, such as a new X-band radar at Shemya at the end of the Aleutian chain. Another year passed before the next flight intercept test, IFT-6, was conducted. The intercept was successful except that the real-time hit assessment performed by the ground-based X-band prototype radar on Kwajalein incorrectly reported the hit as a miss. IFT-7 conducted in early December 2001 also was successful. To this point, all of the GMD flight intercept tests had been essentially the same from the point of view of the target cluster. In these five intercept tests, the target cluster consisted of the target reentry vehicle, a single large balloon, and debris associated with stage separation and decoy deployment. Then in IFT-8, conducted on March 15, 2002, two small balloons were added to the target cluster. This flight intercept test also was successful and marked an important milestone for the GMD program. Two years ago, the objectives of IFT-8 were originally scheduled for IFT-6. However, since IFT-6 and IFT-7 were repeats of earlier tests (IFT-3, 4 and 5), IFT-6 wasn't as complicated as it had once been planned to be, and the objectives of what had been IFT-6, were not achieved until IFT-8, about two years later than planned. So the objectives of IFT-8, originally planned for IFT-6, had slipped roughly two years, or roughly a day-for-day slip rate. Similarly IFT-7 was originally to be the first test with a tumbling reentry vehicle, but as actually conducted IFT-7 was only a repeat of the earlier tests IFT-3, 4, 5 and 6 instead. So tests with tumbling reentry vehicles are still to come. This is not to imply that the program has made no progress at all in the last year or two; but rather that the progress has not been as rapid as planned, nor as rapid as the simple numbering of the tests would imply. By the numbering of the tests you might think that there are only 12 flight intercept tests to go to reach the canonical 20 tests mentioned earlier. In fact there are closer to 15, and there will probably need to be even more. The decision of the MDA to repeat essentially the same test in IFT-3, 4, 5, 6 and 7 was a good thing. They learned a great deal from this series, and it was far better to have repeated those tests for real understanding than to have tried to shortcut ahead without it. If I have one recommendation to Congress, it is to look at the content of the tests, not the numbers. As the years go by it will become more and more difficult for the Congress and the American public to know what's really going on in missile defense if this kind of basic information is not forthcoming from the Missile Defense Agency. No matter what the Pentagon's final missile defense classification and secrecy policy turns out to be, paying close attention to the content of the tests will take persistence and extensive Congressional staff work, and will be vital for the Congress to really understand what is real and what isn't.
An area of content the Congress should watch closely is the role of prior information in missile defense tests and in any system that is actually deployed. Missile defense systems are highly automated, and preprogrammed. This is necessary because the systems either cannot measure, or do not have the time during an engagement to measure, all of the parameters needed for a successful intercept. For an intercept to be successful, prior information is required such as the composition of the target cluster, its trajectories and flight conditions, the look and behavior of the target reentry vehicle, and the time and location of launch. While U.S. intelligence agencies would do their best to obtain advance information about an enemy missile, that information might be incomplete, misleading, or even wrong, much as information since September 11th has been. So far the flight intercept tests have been simple and relatively easy to pre-program. Future tests will be more complicated, and more difficult to program, as an actual ICBM attack would be also. It will be important for the Congress to understand and independently assess whether or not the role of prior information is being realistically handled as the missile defense programs move forward.
A continuing question is whether the GMD system, or any other element of the Administration's layered missile defense program, can be deployed by 2004, or even 2010? In the May issue of Arms Control Today I reviewed the status of each of the various missile defense programs. That full article is also attached to this testimony. For all practical purposes, the only part of the Bush layered NMD system that is "real" is the Ground-based midcourse (GMD) system whose development was a centerpiece of the Clinton administration. It is "real" in the sense that six flight intercept tests have been conducted so far, whereas there have been no flight intercept tests of either the boost or terminal phase intercept systems which are still "paper" systems in most respects. Of course, the Ground-based midcourse system still needs to show it can deal with decoys and with more stressing engagements, but the GMD program does have a beginning. The boost and terminal phase systems - whatever they are or will be - haven't been defined, nor has an overall architecture. As a result the Bush layered system is defaulting to the Clinton Ground-based midcourse system. Furthermore, delays in the THAAD and Navy programs, and in the Airborne Laser, the Space-Based InfraRed satellite systems, and other space-based systems, discussed below, add to the confusion. As difficult as National Missile Defense is, one area that was thought to be relatively "easy" was development of a new, more powerful booster for the interceptor of the GMD system. Also, that new booster was to be incorporated into the continuing series of flight intercept tests, to make those tests more production representative and to be sure that the higher accelerations associated with the new booster did not adversely affect other components or systems on board. However, development of the new booster is about two years behind, and flight intercept tests that were to have used the new booster have already come and gone without it. The only element of a "layered", National Missile Defense system that has begun to be tested in flight intercept tests is the GMD system pursued by the Clinton Administration. Accordingly it is nearly impossible to predict when an integrated, layered, National Missile Defense system with boost, midcourse, and terminal phases might be developed. As noted above, given the most recent pace of testing, some part of the GMD system with marginal capability could be deployed by the end of the decade, and possibly by 2008. However, major components of the GMD system will not be ready until after 2010, nor will boost and terminal phase systems.
The President's Nuclear Posture Review (NPR) calls for deployment of an emergency missile defense capability in the near to mid-term, defined in the NPR as 2003 to 2008. Systems the NPR says might comprise this emergency capability are:
Because of the many delays already suffered, neither the Airborne Laser nor a sea-based missile defense system is likely to be operationally effective by 2003 to 2008. Some components of a rudimentary ground-based midcourse system might be deployed in Alaska, but they will not likely have been demonstrated by operational testing to be effective in the 2003 to 2008 time period either. PAC-3 and THAAD were not designed for a terminal role in National Missile Defense, and are behind schedule in achieving their area and theater defense missions to protect U.S. troops overseas. SBIRS-Low also will not be available in the 2003 to 2008 time scale, and faces large cost and schedule overruns. Key elements of the Administration's missile defense program already have fallen too far behind to be effective in a layered missile defense capability to defend the United States against ICBMs by 2003-2008. Compounding these problems are future delays that are likely to occur in these and related systems, delays that will be inevitable given the still untried technology and extraordinarily demanding systems integration requirements in missile defense.
The Ground-based midcourse system requires some means of tracking enemy missiles in flight. One approach is to use high power radars operating in the X-band, that is, at a frequency of about ten billion cycles per second. These radars have excellent discrimination because of their very short wave length, about a half an inch. A prototype X-band radar on Kwajalein has been part of all of the GMD flight intercept tests so far, and technically, X-band radar progress has been one of the most successful developments in missile defense technology. However, in a deployed system, such radars would need to be placed at different locations around the globe such that the curvature of the earth does not obscure the launch and boost of an enemy ICBM accelerating towards the United States. Locating X-band radars could raise geopolitical issues, as a proposed X-band radar site in England already has. Further, an X-band radar in Alaska cannot see missiles being launched from, say, Iraq, Iran or Libya, and so radar sites would be needed on the east Coast of the United States as well as overseas. A year and a half ago, Lt. General Kadish testified to Congress that establishing an X-band radar on the island of Shemya at the end of the Aleutian chain was the "long pole in the tent." This meant that the X-band radar was on the critical path for an overall GMD system, and that if that radar was not built soon, the GMD system would start slipping day for day. Then as now there were many other developments that would take as long or longer than an X-band radar at Shemya. Nevertheless the Pentagon's official position was that construction needed to start in the Spring of 2001 at the latest. Surprisingly, the Bush Administration has not requested funding for an X-band radar at Shemya in either of its first two budgets. This may be because the Administration has viewed such an installation as inconsistent with the ABM Treaty which the Administration has said it will not violate while the Treaty is still in effect. Also, the Missile Defense Agency has been exploring "portable" deployment of X-band radars, that is, X-band radars mounted on ships, barges or sea-going platforms, or - like the Army THAAD radar - deployable in large airlift aircraft. Portable X-band radars could have geopolitical advantages over fixed, permanent X-band radar sites. Instead, or at least until the ABM Treaty has been abrogated on Thursday, June 13, 2002, the Administration has been emphasizing a plan to upgrade an existing radar at Shemya called "Cobra Dane". That can change come Friday. Under the plan so far, the Cobra Dane radar on Shemya, (which originally was a perimeter acquisition radar (PAR) for an ABM site in the United States in the early 1970s but then deployed in Alaska to track Soviet missile tests), would become an advanced early warning radar with some target cluster discrimination capability. However, Cobra Dane is a fixed radar pointed towards the Kamchatka Peninsula and cannot see test launches from elsewhere in Alaska, from California, or from Kwajalein. Thus, Russia might have viewed the upgraded Cobra Dane radar as a deployment violation of the ABM treaty, not simply as a test facility, since the radar cannot see such test launches. In any case, the Cobra Dane radar operates in the L-band with about eight times poorer discrimination than a new X-band radar would have. Overall, an X-band radar at Shemya has been delayed about a year and a half, and deployments in other locations around the world seem to be on hold.
Another important element of the Administration's plan is the Space-Based Infrared Satellite (SBIRS) Program which is to consist of two sets of orbiting sensor satellites, SBIRS-High and SBIRS-Low. SBIRS-High and SBIRS-Low are proposed systems for detecting the launch of enemy ballistic missiles and for tracking and characterizing enemy missiles in flight in support of U.S. missile defense. Some SBIRS advocates believe it could take the place of the X-band radar for missile tracking and discrimination, thereby avoiding the physical limitations of X-band radars and the need to locate them in a number of places around the world. However, SBIRS has its own significant technical issues, and is so far behind schedule that the Administration will probably decide not to put all its eggs in that basket. SBIRS-High is to replace the existing Defense Support Program satellites, which watch for missile launches and provide early warning. SBIRS-High is to have four satellites in geo-synchronous orbit and two satellites in highly elliptical orbits. A year ago, those satellites were scheduled for launch in 2004 and 2006 respectively, but recently those dates have slipped roughly two years. A year ago, realistic operational testing was scheduled for 2007, now it may not occur this decade, which means that full deployment may not occur this decade also. SBIRS-Low is to consist of about 24 cross-linked low earth orbit satellites. A year ago the launch of the first of those satellites was scheduled for 2006 also, but SBIRS-Low has slipped two years as well. The developmental testing program for SBIRS-Low is very challenging, and realistic operational testing with threat representative targets probably will not begin this decade either. This in turn could delay the full constellation of SBIRS-Low satellites until the middle of the next decade. Costs for SBIRS-High are overrunning and are in danger of a breach of the legal limits for defense acquisition cost growth. Pentagon officials are examining ways to restructure the program. SBIRS-Low also is over budget, but much more dramatically with the latest reported cost estimate being at least $23 billion, not the $10 billion estimated just last year. In the meantime, the twenty-five year old Defense Support Program satellites provide good missile warning, but not the discrimination and characterization capabilities desired from SBIRS.
The Administration has said it will deploy by 2004 a rudimentary ground-based midcourse system consisting of a small number of interceptors taken from the test program, and coupled with an upgraded Cobra Dane radar. As described above, by 2004, a residual capability constructed with test assets in Alaska would need to rely on the Cobra Dane radar, since an X-band radar radar will not be in place. Since the Cobra Dane radar is a fixed radar pointed towards Kamchatka, it cannot see test launches from California, Kwajalein or Kodiak. As a result the MDA must confront the question as to when if ever they will test the residual system with target missiles fired from near the Kamchatka Peninsula, perhaps dropped from aircraft where the Cobra Dane radar could see them, and toward the planned missile silos near Fort Greely in Alaska. Coming as soon as 2004, the residual capability would also have to rely on interceptor missiles that will not have been tested under realistic operational conditions, since it will not be possible to conduct 15 or 20 developmental tests and a handful of realistic operational tests in the next two years. Also this residual capability would not have the intended capabilities of the SBIRS satellites because of the difficulties that program has been having. As a result, the residual capability would not work against missiles fired from Iraq, Iran, Libya, China or Russia. Against missiles fired from North Korea, the residual system would only have been demonstrated to work if those missiles flew on trajectories and flight conditions that were known in advance and similar to those already tried in earlier flight intercept tests, within range of the Cobra Dane radar, and with very simple decoys or no decoys at all. Perhaps someone could say this would be better than nothing at all, but the money would be better spent on systems such as we saw demonstrated in Afghanistan that could detect and blow up a North Korean missile on the ground before it was fully assembled and launched. Perhaps most important, the proposed residual capability would not prevent or deter an enemy from threatening our friends and allies. This is a central tenet of the Administration's approach to missile defense, but it would not be achieved with this residual capability. Deployment of a missile defense system means the fielding of an operational system with some military utility which is effective and suitable under realistic combat conditions, against realistic threats and countermeasures, possibly without adequate prior knowledge of the target cluster composition, timing, trajectory or direction, and when operated by military personnel at all times of the day or night and in all weather. These operational considerations will become an increasingly important part of test and simulation plans over the coming years, Mr. Chairman, that completes my prepared remarks. I would be pleased to take your questions.
 MISSILE DEFENSE HOME CDI HOME
1779 Massachusetts Ave, NW, Washington, DC 20036-2109 Ph: (202) 332-0600 ยท Fax: (202) 462-4559 info@cdi.org |
