Update on

Missile Defense Technology

W A S H I N G T O N R O U N D T A B L EO N S C I E N C E & P U B L I C P O L I C Y

By Gregory Canavan

Washington, D.C.

The George C. Marshall Institute

The George C. Marshall Institute, a nonprofit research groupfounded in 1984, is dedicated to fostering and preserving the integrity ofscience in the policy process. The Institute conducts technical assessmentsof scientific developments with a major impact on public policy and com-municates the results of its analyses to the press, Congress and the public inclear, readily understandable language. The Institute differs from otherthink tanks in its exclusive focus on areas of scientific importance, as well asa Board whose composition reflects a high level of scientific credibility andtechnical expertise. Its emphasis is public policy and national security issuesprimarily involving the physical sciences, in particular the areas of missiledefense and global climate change.

The Washington Roundtable

on Science and Public Policy

The Washington Roundtable on Science and Public policy is aprogram of the George C. Marshall Institute. The Roundtable examinesscientific questions that have a significant impact on public policy and seeksto enhance the quality of the debate on the growing number of policydecisions that look to science for their resolution.

The opinions expressed during Roundtable discussions do notnecessarily represent those of the Marshall Institute or its Board ofDirectors. Additional copies of this transcript may be ordered by sending$7.00 postage paid to:

The George Marshall Institute1625 K Street, NW - Suite 1050

Washington, D.C. 20006Phone: 202/296-9655

Fax: 202/296-9714E-mail: info@marshall.org

Website: www.marshall.org

UPDATE ON MISSILE DEFENSE TECHNOLOGY

Gregory H. CanavanMay 22, 2002

The George Marshall InstituteWashington, D.C.

Gregory Canavan has worked in the area of missile and defensetechnologies for over twenty-five years. He has served as Director of theOffice of Inertial Fusion at the Department of Energy and as Deputy to theAir Force Chief of Staff. Since 1981, Dr. Canavan has been ScientificAdvisor, Physics Division of Los Alamos National Laboratory. In 2000, hewas elected a Fellow of the American Physical Society for his contributionsin military science and technology and their transfer to the civilian sector.

UPDATE ON MISSILE DEFENSE TECHNOLOGY *

Gregory H. CanavanMay 22, 2002

SummaryThis note discusses developments in missile defense technology and

the prognosis for robust protection based on them. The competitionbetween the offense and defense is unlikely to be settled by a stroke, andwould not remain settled if it was, but current technologies and systemscould significantly reduce the utility and attractiveness of missiles to roguesand would-be proliferators. The current program would serve as a first stepin a continuing spiral of conflict between offenses and defenses. The appro-priate goal is to put missile defenses so far ahead of offenses that theywould dissuade others from engaging in missile competitions altogether,which is not beyond the capability of near term missile defenses that use allavailable layers and concepts.

Dr. Canavan’s RemarksThreat. The number of nations with missiles grew from six in 1972

to 16 in 2001. SCUDs spread to the Middle East, Asia, and other parts ofthe world. In 1998 and 1999 Pakistan and India tested their 1,300 kmGhauri and Shahab-3 missiles, India tested its 2,000 km Agni II MRBM,China prepared to deploy its 8,000 km range mobile DF-31, North Koreatested its Taepo Dong 1, and developed the Taepo Dong 2, which haslarger payloads.1 China is helping Pakistan, Iran, Libya, and North Koreawith their missile programs, and North Korea is helping countries of theMiddle East, South Asia, and North Africa.

The CIA states that countermeasures to defensive sensors that arenow “readily available technology” include “separating RVs, spin-stabilizedRVs, RV reorientation, radar absorbing material, booster fragmentation,jammers, chaff, and simple balloon decoys.” Separating an RV permitsthe attacker to distance it from its booster and orient it to reduce its signa-ture. Spin stabilizing permits reduces signatures, eases reentry, and increasesaccuracy. Absorbing material reduces radar returns and range. Frag-mentation, chaff sized to the radar frequency, and balloon decoys increase the

1

* The views expressed by the author are solely those of the author and may notrepresent those of any institution with which the author is affiliated.

number of “traffic” decoys a radar must examine. Jammers deny the radar-information about RV range and angular location. That these technologiesare “readily available” means that techniques and systems developedover decades at considerable expense by the US and USSR are now avail-able at a fraction of that cost to rogues.

Ballistic missiles launched from the rogue nation are not the onlyvehicles available to them. Cruise missiles, short-range missiles, or surfacetransport are also attractive options. However, ballistic missiles are thevehicle chosen by a number of nations. As long as there are no defensesagainst them, they will remain attractive if expensive modes because oftheir certainty of delivery, and the lack of defenses against them will reducethe incentives for closing other modes of delivery. Conversely, progress onballistic missile defenses would increase the incentives for closing othermodes.

Epochs. Missile defense efforts can be divided into four eras:nuclear systems, interim technologies, the SDI-GPALS deployments, and theTMD-NMD limits.2

2

Trends in Missile Threats

• Nations w/ missiles: six in 1972 16 in 2001– SCUDs to the Middle East, Asia, etc.

– Pakistan tested 1,300 km Ghauri and India Shahab-3 and 2,000 km Agni II

– China ready to deploy 8,000 km range mobile DF-31

– North Korea tested Taepo Dong 1 and developed Taepo Dong 2

• China helping Pakistan, Iran, Libya, and North Korea– North Korea helping Middle East, South Asia, and North Africa

• Missile and RV technologies are now widely available. – Countermeasures to defenses now “readily available” include “separating

RVs, spin-stabilized RVs, RV reorientation, radar absorbing material, booster fragmentation, jammers, chaff, and simple balloon decoys.”—CIA

• Rogues have chosen missiles

Figure 1: Trends in Missile Threats

Nuclear. The NIKE, Sentinel, and Safeguard systems were based onradars and nuclear interceptors. NIKE and Sentinel were directed towardslight attacks on population, for which they were well suited. Spartan was apolitically dictated defense of the missile deterrent with the Sentinel tech-nology, for which it was ill suited. NIKE was compromised by its soft radars.Sentinel addressed that by the development of phased array radars.However, Safeguard’s PAR radars could not operate reliably in the nuclearenvironment produced by its Spartan interceptors and offensive bursts.

The US concluded that ABM systems based on radars and nuclearinterceptors could not work and convinced the USSR to abandon attemptsto defend territory and populations by them or any other technologies inthe ABM Treaty.

Interim developments between the Treaty and the SDI were inthree major areas. LoADS was a nuclear endo-atmospheric system. Itanswered the criticisms of Safeguard for defense of the deterrent but wasrejected for Treaty concerns. DARPA initiated the scaling of directed energy(DE) Space Based Lasers (SBL) to the levels required for BMD, but foundthat countermeasures and the vulnerability of the platforms could reduce its

3

Program object type elements weakness

NIKE/Sentinel pop. nuc int, radars endo radarsSafeguard deter nuc int, radars exo radars

••

AMB Treaty ABM all & OPP progressInterim– LoADS deter nuc endo int, radars Treaty

– DARPA tech DE SBL vulnerability,CM– Army tech KE HOE, ERIS cost

• SDI deter DE/KE SBI, SBL… costGPALS protect KE BP, GBI, BE Treaty

••

•TMD troops KE PAC-3,THAAD cost, scheduleNMD pop KE GBI,GBR,EWR cost, discrim, bkgnd

•••

Figure 2: Brief History of MD Technology

effectiveness. The Army developed kinetic energy (KE) interceptors anddemonstrated their ability to perform non-nuclear hit to kill intercepts,although the early versions were heavy and expensive.

Strategic Defense Initiative (SDI) supported research on a widerange of DE and KE concepts. Its goal ultimately became improving deter-rence because of the difficulty and cost of its technologies. The vulnerabilityof DE and space based interceptors (SBI) led to systems that were thoughtto be unaffordable. Global protection against Limited Strikes (GPALS) hada goal of affordable protection of the US and its allies against limited strikesfrom anywhere on the globe. It used a Brilliant Pebble (BP) boost phaselayer and a KE midcourse layer. The resulting system was evaluated to sur-vive determined suppression attacks, achieve adequate protection, and costan order of magnitude less than previous systems, but it was in conflict withthe Treaty.

Theater missile defense reduced development to threats todeployed forces. The North Korean Taepo Dong launch forced the resump-tion of National Missile Defense (NMD), although the version chosen waslimited to a single site and earlier technology.

4

• Objective: high confidence of protection – whole US and allies – from limited strikes (~ 1 SSBN)– Independent of launch point

• Dominant attrition by Brilliant Pebble (BP)– Small (few kg) autonomous space based interceptor– several BP/missile for high Pk

• Remainder by KE underlay – Ground Based Interceptor (GBI)– Supported by Brilliant Eye (BE)—later GBR/UEWR

Figure 3: GPALS Performance Issues

Systems with military objectives — NIKE, Sentinel, LoADS, TMD,and GPALS — could have met their objectives, had they been pursued totheir conclusion. Those which had politically directed objectives — Safe-guard, SDI, and NMD — could not. The ABM Treaty was effectively ananti-technology program, which fell into the second category and failedbecause technological progress invalidated its basic assumptions.

The nuclear epoch developed most of the radars used today,although nuclear interceptors have not been developed in recent decades.Current KE and DE concepts were developed during the Interim. SDIadded layered systems and command and control (C2). GPALS used BPand the ground based interceptor (GBI) to provide robust protection.Brilliant Eye satellites (BE) were to have tracked buses and RVs to performthe functions earlier performed by radars. TMD was a retreat from defenseof population.

NMD was a limited defense against small, simple attacks. BecauseBP had been eliminated, GBI was the only intercept layer. BE was delayedand DSP/SBIRS-High could not provide intercept-quality trajectories, soNMD had to use radars for detection and discrimination. The x-band radar(XBR) could discriminate, but was not suitable for search, so upgraded earlywarning radars (UEWR) developed in the nuclear era had to be used, whichreintroduced their known vulnerabilities into the current system.

Layered defenses. Terminal, midcourse, and boost phasedefenses each have strengths and weaknesses. Terminal defenses can pro-vide additional protection cheaply for high value targets, but their defendedfootprints are small for fundamental reasons. Midcourse defenses have themost space and time to act, but are most sensitive to decoys. Boost phasedefenses are insensitive to decoys, but must act very fast. Boost phasedefenses from the surface or air are probably preferred for threats thatpermit ready access; space-based defenses for extended or multiplethreats. Effective multi-layer defenses minimize the countermeasures opento adversary as well as the options for overloading any given layer.

5

Performance Issues. Nuclear interceptors performed well, as didendo-atmospheric radars. Exo-atmospheric radars worked adequately onlyfor small threats. The DSB concluded LoADS could defend Minuteman bet-ter than multiple shelter concepts. SDI had limited performance for accept-able cost, amplified by vulnerability of SBL and SBI, which made defense ofthe deterrent impractical.

Figure 5 shows the number of RVs killed in boost, post-boost, andmidcourse for the ICBM portion of the SDI threat, the launch of Russia’sSS-18s ICBMs, 300 missiles with 10 RVs each, from its distributed launcharea. In these illustrative calculations for a N = 1,000 constellation ofboost phase interceptors, ≈ 500 RVs are killed in boost, 400 in post boost,500 in midcourse, and 1,100 leak through all layers. For N = 2,000 thenumber of kills are ≈ 1,000 in boost, 700 in post boost, and 200 in mid-course with little leakage. Distributed launches made it possible for a largefraction of the SBI to participate in the boost or post boost phases, whichreduced absenteeism to ≈ 1,000 SBI/1,500 RV ≈ 7 SBI/SS-18.

6

0

500

1000

1500

2000

2500

3000

3500

0 500 1000 1500 2000 2500

N

kill

s

Kboost

Kbus

Kbus&boost

Leak bus&boost

Kmidcourse

Ktotal

Figure 5:Heavy ICBM RV kills and leakage as functions of

interceptor constellation size

After the fall of the Soviet Union, GPALS’s goal shifted to high-confidence protection of population of the whole of the US and its alliesagainst limited strikes from any launch point. The dominant attrition was tobe provided by BP, which would require several BP per missile for a highcompound kill probability.3 The remainder of the attrition was to be pro-vided by KE underlay consisting of the ground based interceptor (GBI)supported by the Brilliant Eye (BE) satellites derived from the BP. WhenBE was delayed, it was necessary to reintroduce the GBR and UEWR.The GPALS ICBM threat was about 10 missiles. Fig. 5 scales linearly onthe threat in that region, so ≈ 2,000 (10/300) ≈ 70 BP should provide therequired coverage, with absenteeism as above.

Figure 6 shows the launch of all 20 missiles with 10 RVs each froma single SSBN, which is a more stressing target because it is a point sourceof missiles, which tends to poke a hole in the defensive constellation. AtN = 120, boost phase kills are reduced to ≈ 20 RVs, although post boostincreases to 100. Midcourse contributes ≈ 60 and ≈ 20 leak. IncreasingN to 180 would reduce leakage to low levels with over 90% of the kills inthe boost or bus phases. Because of the concentration in space and timepossible with such SSBN launches, they were the most stressing threats forboth SDI and GPALS and required similar constellation sizes for both.

7

Figure 7 shows performance against the still more stressing SS-25mobile missile, which only accelerated for ≈ 150 s and had 1 RV per mis-sile so there was no bus phase. Even for a constellation of N = 450 thereare only ≈ 12 RV kills in boost. As constellation sizes grow, midcourse killsare slowly replaced by boost phase kills. As the SS-25 was ill suited to usemidcourse decoys, the numbers of midcourse kills shown are appropriate.The SS-27 has both a short burn and few RVs and a bust from which todeploy decoys.

8

0

50

100

150

200

250

0 20 40 60 80 100 120 140 160 180 200

N

kill

s

Kboost

Kbus

Kbus&boost

Leak bus&boost

Kmidcourse

Ktotal

Figure 6:SLBM RV kills in each phase as functions of constellation size

The SBI and BP technologies underwent 7 major reviews by theDSB, JASON, American Physical Society, and others in 1990, which sup-ported BP’s passage through the Defense Acquisition Board and establish-ment as a Major Defense Acquisition Program. Those reviews established aCAIG cost estimate of ≈ $10 billion for development, deployment, and lifecycle operations with a cost estimate of $1 million per BP on orbit,4 whichis contested, but remains the only reviewed estimate.

As an MDAP, BP went through testing comparable to that for themidcourse NMD system; however, it was reduced to a “robust” technologyprogram 1991 and abolished altogether in 1993 for non-technical reasons.A nucleus of researchers in national laboratories and industry used residualsensors to execute the 1994 Clementine mission, which remapped themoon at high spatial and spectral resolution, and the Astrid demonstration,which flew efficient pump fed engines of the size required for BP.5

TMD and NMD. The Gulf War showed the need for TMD. Its develop-ment was started during GPALS. The Clinton administration emphasizedTMD but cut its funding ≈ 25% and allowed its elements to develop

9

0

20

40

60

80

100

120

0 50 100 150 200 250 300 350 400 450 500

N

kill

s

Kboost

Kbus

Kbus&boost

Leak bus&boost

Kmidcourse

Ktotal

Figure 7:SS-25 kills for boost and midcourse phases vs constellation size

stovepipes. It disestablished Global defenses and reduced NMD to a tech-nology program. The Rumsfeld Commission was established to see whethersuch actions were consistent with threats. It concluded that missile threatswere becoming less predictable in scope and timing and that a capabilitybased approach could be more appropriate, which was emphasized byNorth Korea’s launch of a three stage Taepo Dong over Japan, which wasprobably a space launch but could as well have been an ICBM. That forcedthe resumption of NMD, but the version chosen was nominally Treatylimited and consisted of a single midcourse layer.

NMD’s key elements were the DSP and SBIRS -High and –Lowsatellites, UEWR search radars, and XBR (GBR); the GBI Interceptors; andbattle management, command, control, and communications (BMC3). Thesatellites and radars were inherited from much earlier programs. GBI wasintended for the midcourse layer for GPALS, so it had technology fromsomewhat before 1990. The BMC3 developed was essentially a fire controlsystem for GBI range tests.

10

• 7 major reviews (DSB, JASON, APS, etc.)• Defense Acq Board — Major Defense Acq Prog

– Cost ~ $10B w/ $1M/BP on orbit

• Testing ~ midcourse NMD• Reduced to “robust” technology program 1991

– abolished in 1993 for non-technical reasons

• Nucleus in national laboratories & industry– executed Clementine, Astrid demos in 1994

Figure 8: Interceptor Technology

Development in SDI & GPALS

11

• Gulf War showed need for TMD– TMD started during GPALS

• Clinton administration cut 25% & splintered– disestablished NMD, Global defenses

• Rumsfeld Commission• North Korea Taepo Dong Launch• Resumption of NMD

– Treaty limited– Midcourse only

Figure 9: Transition to TMD & NMD

Sensors:– DSP/SBIRS-High & -Low

– UEWRS

– GBR/XBR

Interceptor– GBI

BMC3

•

•

•

Figure 10: National Missile Defense Elements

Results of IFTs. GBI’s half dozen integrated flight tests (IFTs) onthe Pacific range are too small to form statistically meaningful conclusions.IFTs have been criticized in that auxiliary elements such as beacons andGPS have been used to simulate elements that are not yet available orimprove range safety, illumination conditions have been repeated, andgeometries have been restricted by Treaty limitations. IFT-1A and –2 suc-cessfully tested GBI EKV sensors performance against representative lightballoon decoys and IFT-3 performed a successful intercept in their pres-ence. IFT-4 and –5 failed because of cooler, booster, and quality controlproblems rather than fundamental issues. These problems kept NMD frombeing ready for deployment. IFT-6 and –7 produced successful intercepts.

Specific countermeasures have been proposed for test. Balloonswere ineffective in IFT-1A and –2. Super-cooled shrouds are technicallyinteresting, but probably not within the indigenous capabilities of roguenations. Sub-munitions are effective against Treaty-limited interceptors, butcan be engaged effectively by multiple KV interceptors. The full impact ofanti-simulating RVs to resemble decoys has not been assessed.

12

• Results of IFT-1A, -2, -3 / -4, -5 / -6, -7,…• Criticism of tests: elements, geometry, CM• Specific Countermeasures

– Balloons

– Shrouds

– Submunitions

• Antisimulation

Figure 11: NMD Test Results

Boost Phase Intercept Options. A number of rogue threatscould permit access to ICBMs in their boost phase from surface-based boostphase systems. Such intercepts would be effective if the missile’s trajectoryallows access to boost from ships in international waters or secure landbases, as would be the case for many North Korean launches toward theUS. As launch areas are moved further inland, surface based interceptorsrequire higher velocities and accelerations and shorter release times. Whenthere are no secure bases with useful access or there are a number ofwidely separated launch areas, absenteeism is less of a penalty, so SBI orBP are preferred over surface based boost phase intercept systems.

Figure 13 shows the distances flown by the interceptor and theinterceptor and target for head on intercepts of ICBMs by 4 and 7 g inter-ceptors, whose ideal ranges are reduced by 30-50% by typical operationalrelease delays.6 Standoff distances are adequate for some launch areas suchas North Korean trajectories over water, but not for those further inland.

13

• Surface-based boost phase systems preferred where trajectories allow access to boost from – Ships in international waters – Secure land bases– Require high velocity, acceleration, response

• Space based interceptor (SBI) preferred– No secure bases with useful access– Numerous, separated launch areas

Figure 12: Boost Phase Intercept Options

14

0

500

1000

1500

2000

2500

0 10 20 30 40 50 60 70

delay (sec)

rang

e (k

m) GBI dist 4g

Total range 4g

GBI dist 7g

Total range 7g

Figure 13:Flyout distance and range for fast ground based interceptor vs delay time

0

0.05

0.1

0.15

0.2

0.25

0 2 4 6 8 10 12

Vmax (km/s)

Co

st (

$B)

a=6g

a=12g

a=24g

Figure 14:Constellation cost vs Vmax for 240 sec rogue ICBM

(space segment procurement only)

Figure 14 shows the cost for SBI to intercept a single rogue ICBM.It has a minimum at ≈ 6 km/s for 6 g, which shifts to ≈ 8 km/s for higherSBI accelerations. At 6 g the minimum is ≈ $0.1 billion/missile, so two SBIon five simultaneously launched missiles would cost ≈ $1 billion to purchasethe on-orbit hardware.

Theater and regional missiles are more stressing in that theirboost and flight times are shorter, but they can make less effective use ofdecoys and countermeasures.

Figure 15 shows the ranges from which SBI can fly in to interceptin boost, which are 500-700 km for 3,000 - 5,000 km regional missilesbut fall to a few tens of km for 500 - 1,000 km theater missiles.

Figure 16 shows the resulting RV kills as a function of the range of5 simultaneously launched ICBMs for N = 500 and 1,000 SBI constella-tions. For regional ranges, boost phase kills are significant and absenteeratios approach those for ICBMs. Below 2,000 km most RVs are killed inmidcourse. Below 1,000 km almost all kills are in ascent or midcourse and

15

0

100

200

300

400

500

600

700

800

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

theater missile range (km)

inte

rcep

tor

ran

ge

(km

)

a=6

a=12

a=24

Figure 15:Interceptor range as a function of theater missile range

(point launch of 4g theater missile, 30 sec interceptor delay)

some RVs leak through. As constellation size increases, kills are slowlyshifted from midcourse to boost.

The Current Program is to produce defenses in all three layersASAP from existing elements. In the Terminal phase that means a two-layer system based on THAAD and Patriot, with allied contributions asappropriate. Regional systems do not now exist but are needed to defendallies and friends. They can be based on THAAD derivatives. The likely ele-ments of the National defense remain DSP, UEWR, and GBI. Boost phaseelements include ship and air based interceptors and space experiments.

It is now recognized that the BMC3 to integrate the various layers isof equal importance.

The space-based interceptor is a natural complement to these ele-ments. It has its own detection and track sensors, so DSP and SBIRS arenot essential for engaging ICBMs, although it could be useful for Theaterand Regional intercepts in boost. Improved performance is not necessaryfor rogue ICBMs; previously developed configurations could be resumed,updated, and tested. The space based interceptor program was about half

16

0

1

2

3

4

5

6

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

range (km)

kill

s

Kboost N=500

Kboost N=1000

Kmid N=500

Kmid N=1000

Ktot N=500

Ktot N=1000

Figure 16:Number of RV kills as a function of TBM range

(6 km/s, 25 g BP)

way through when cancelled for non-technical reasons, so it should be pos-sible to finish testing in 2-3 years, the time scale for ground-based ele-ments. The two major elements are an improved engine, which is notneeded for ICBMs, and a demonstration of plume to missile handover,which was difficult to test under the Treaty, but could be started with exist-ing sensors packages. The previous DAB estimates would give a cost forthe few hundred interceptors needed for coverage of near term ICBMthreats of ≈ $3-5B, which would also cover most regional missiles.

Summary. The last four decades have produced a number ofobservations that seem robust. Radars could not address large exo-atmos-pheric threats, but KE sensors apparently could—if supported by some-thing other than radars. DE should be useful on a longer time scale forfaster missiles. TMD alone did not cover all emerging threats; it seemed tostimulate development of regional and ICMB threats. NMD based onground-based midcourse alone was adequate for small, simple threats, butprovides expensive coverage and lacks robustness. Space based intercep-tors could apparently provide affordable protection against them as well asaccidental and unauthorized launches.

17

• Most MD systems could have met original objectives

• Radars could not address large exo threats– KE early — DE for faster missiles, longer term

• GPALS survive & protect with developed layers– Affordable for large threats

• TMD did not cover emerging threats

• NMD coverage of small, unsophisticated threats– expensive & lacked robustness

Figure 18: Summary

Conclusions. The Missile Defense Agency (MDA) has as its goalthe development and deployment of effective defenses in each possiblelayer for the US and its Allies as soon as possible using existing technologyand systems, which are to be upgraded in a spiral through block modifica-tions. The technologies discussed above could make effective defenses pos-sible in each layer on the time scales desired. The scope of MDA iscomparable to that of previous national programs, but they had definitegoals, short time scales, strong leadership, multiple options, and strongR&D. MDA has an open ended program designed by a committee on thebasis of a subset of the technologies available. Spiral development cannotovercome intrinsic problems by deploying more of the same or similar systems.

Space based interceptors were half way through developmentbefore they were cancelled for political reasons, which suggests that theirdevelopment could be completed on roughly the time scale as other ele-ments of the current program. Doing so would produce a combined defensewith affordable boost and midcourse layers.

The decadal or millennial competition between the offense anddefense is unlikely to be settled by a stroke, nor is any solution likely toremain static. There was a potential solution in GPALS that could have

18

• Current elements could provide protection– Not large, accidental, or unauthorized

– Program is narrow — little R&D, few options

– designed by a committee

• Space elements essential for robust protection

• Goal should be to put missile threats out of reach– A balanced program to do that must use all technologies

Figure 19: Conclusions

largely eliminated the utility of offensive missiles, but it was lost during theClinton administration in deference to the Treaty. The MDA programwould serve as a first step in a continuing spiral of conflict between offensesand defenses. The appropriate goal is to put missile defenses so far aheadof offenses that they would dissuade rogues and others from engaging inmissile competitions altogether. It is not beyond the capability of missiledefenses, even those attainable in the near term. The tools are at hand,but only a portion are being developed. A balanced program must developand use all.

# # #

Questions and answers.

Q: When do you believe we will see missile defense really in action aroundthe world? What kind of defense will it be?

Canavan: Actually we did — we saw Patriots fired in the Gulf War.

Q: But those were not really anti-ballistic missiles.

Canavan: Yes, they were anti-ballistic missiles. SCUDs were ballistic mis-siles. They weren’t very good missiles, but they actually turned out to befairly sophisticated targets as they broke up during reentry. You can argueabout the technical performance of the sensors of the Patriot missile, butthere is no question that having the Patriot in operation kept Israel out ofthe war. If they had gotten into the war, the coalition would have evapo-rated. So we’ve seen missile defense in action. The question is, when andwhere are we going to see it next? That’s a harder business. I don’t knowwhere for sure we are going to see it, but I would be willing to bet that weare going to see it in the next decade.

Q: What kind of missile defense do India and Pakistan have?

Canavan: None. India and Pakistan are basically at the level of the UnitedStates and Russia at the time of the Cuban missile crisis. A lot of offensivemissiles and no defenses whatsoever. Thus, their behavior is similar.

Q: Why the focus on kinetic kill mechanisms instead of an explosive?

Canavan: I mentioned the HOE experiment as the first demonstration thatkinetic energy actually worked. HOE actually had a several meter kill

19

enhancer. From time to time people have used explosives. The Arrow, forinstance, uses an explosive kill enhancer instead of just a warhead. I haveno strong preference one way or the other. I would say in support of kineticenergy weapons, that when they hit, they really hit. In situations wheretheoretically you would expect to miss by tens of centimeters, as far as youcan tell they have hit by centimeters. I believe in the value of a killenhancer if the miss distance is larger than the target. Brilliant Pebble hadhigher accuracy, so it did not need one. I am not sure that whether or notto use a kill enhancer is a critical issue with today’s accuracy.

Q: Do you imagine that would be possible to engage the SS-25 missile inboost phase?

Canavan: Let me parse the question. When you have a point launch of SS-25s deployed out of garrison, you can draw a circle around the missilewhose radius is the interceptor velocity times the boost phase duration. Thearea from which interceptors can reach boost is the square of this radius,so as the boost phase gets shorter, a smaller fraction of the interceptors inorbit can reach the missile in boost.

Q: It can reach them afterwards.

Canavan: Yes, I am trying to parse the question. The previous commentsconcerned boost. Missiles can be intercept later in the ascent or bus phase,but there are additional issues there.

Q: Maybe it’s just a semantic thing, I mean, boost phase is ten minutes; aslong as it cannot take evasive actions, you can shoot it down. So it’s not aproblem.

Canavan: I agree that you can engage the missile in later phases with theright sensors, but they are different than those used for boost phase.

Q: So why do you say it’s a problem?

Canavan: The 22-25s boost phase is very short, about 125 sec. It has nobus and hence no bus phase. However, the SS-25 has no bus, so it cannotdeploy decoys. Thus, the reentry vehicle (RV) remains an attractive targetbecause it is not concealed in decoys. Other Russian missiles do have busesand can deploy decoys relatively quickly. The new SS-27 has a fast missileand bus and decoys, so none of these factors aid the defense.

20

Q: Well, multiple RVs as you define them is where the problem begins, withthe SS-27. But as long as [the interceptors] catch [the RVs] before then,the ballistic trajectory, one should probably find it more like when you candeploy decoys as the terminal phase of boost phase intercept missions.Otherwise you are missing the key point. So I really think it’s a bad choiceof words, which has awful consequences.

Canavan: I stand corrected, but let me draw a distinction in technology. Ittakes a different set of sensors to look at a missile in boost than it does tolook at a RV post-separation. It’s important to keep that distinction inmind, which is why I am attempting to differentiate between intercepts inboost, post-boost and mid-boost. Once the RV is separated, which happensas soon as the SS-25 burns out, you’re only the long wavelength infraredsignature to go on, which requires a different sensor array.

Q: Before giving up on the point, since the vehicle is not doing anything,especially when it takes up to eight minutes, we don’t know where the vehi-cle is until later; it comes in perfectly. So again, I say there is no differencein what you and I advise. Eight minutes, ten minutes, or whatever, youhaven’t got this decoy capability. Until then, the world is fairly simple; youhave an easy job of shooting it down and you can do it with all types ofdefenses without resorting to something that’s really complicated.

Canavan: I agree. That’s an excellent point. For an SS-25, knowledge ofits trajectory up to burnout could help the search for it in post-boost.However, trajectory measurements in the boost phase are not sufficientlyaccurate to fire interceptors open loop there or in midcourse or terminalintercept layers.

Q: Although this is really very central point. I am not sure what you’reagreeing to, because Theodore Postol and Richard Garwin argued thatshortly after 31/2 minutes of boost, up to maybe 4 minutes, that the RV willseparate and at that point, decoys and countermeasures can be deployed,so that in midcourse you have the discrimination problem, which is thebasis of their entire opposition to midcourse defenses. It’s ironic that theirfriends on Capitol Hill spend their time cutting out the boost-phase moneyfor sea and space, which is designed to correct that. Postol then e-mailedme and said: “Well, they’re not my friends.” And I emailed back and said:“Well, you brief them supposedly that this is the answer.” So what pointare you talking about between boost and when the decoy, balloons andstuff are deployed?

21

Canavan: I’m sorry, we are talking about two different points. You arereferring to the point of separation of an ordinary missile from its bus. Thatmarks the transition from the boost to bus phase, in which the bus deploysmultiple RVs, balloons, decoy, and other countermeasures. There is animportant distinction from un-bussed missiles, which Baker Spring and Iexplored in in a paper for the Heritage Foundation. Missiles without busesprovide a much simpler target. If you have an un-bussed SS-25 missile,which accelerates for 125 seconds and then releases one RV, burning thatfast will not make it survivable. The reason is that the sensors on a space-based interceptor have to be able to look through frequencies from visibleto LWIR to close on and handover targets. Thus, they are as capable of see-ing a cold RV as of seeing a hot booster. Thus, to a space-based intercep-tor, an SS-25 is just as attractive post-boost as it is pre-boost. It is animportant point because most simple rogue and theater ICBMs are of thatnature.

The other point is just the standard argument, that if a missile witha bus deploys a number of decoys at the outset of bussing, then intercep-tion of the RV with standard mid-course interceptors gets a lot harder atthat point. That is a different problem. Both are interesting. I am sorry ifI’ve confused the two.

Q: To follow up on that, it is easier to decoy after you are up in space andyou get into target problems and accuracy as you do the other way imme-diately after boost. So again, I don’t think any sophisticated attacker woulddecoy early in the first eight minutes, and probably more likely in true mid-course. It simply prolongs your new boost phase.

Canavan: I agree, although we have seen both early and late bussing.Deploying defenses that can intercept in the boost and bus phases putspressure on the attacker to deploy earlier, which imposes a penalty for thereasons you mention.

Q: Some of us agree with the way you stated it in your briefing, rather thantrying to hustle or confuse the boost phase and post-boost. You are talkingabout the SS-27: you can bus immediately but counter-measures are reallyafter the end of busing.

Canavan: As you point out, an inexperienced attacker would probably taketime to deploy decoys, particularly if he intended to decoy the full a rangeof axes. It was then realized that this gave the defense additional leverage.Even for a rogue missile with a single RV, killing the bus before it completesthe deployment of all decoy clusters can reduce the number of clusters that

22

have to be attacked, which reduces the number of GBIs that have to beexpended per missile.

Q: Most of the threats that we should be concerned about are from primi-tives; even China today does not have decoys. They have developed them,sure, but I mean you’re talking about the people the British should be wor-ried about, and they’re not; why should we care about China and other civ-ilized nuclear powers? And for that matter, I think, boost phase, you mayfind it may actually even work on midcourse.

Canavan: What you’re saying is that even if they did deploy decoys, itmight be late in the trajectory, and they might all be in a single cluster.There is no incentive for any attacker to do either.

Q: Could you talk a little about what effect you think might come aboutbecause of the US leaving the ABM Treaty? Are there any technologicalchanges that you foresee?

Canavan: To take an obvious case, you can’t do anything in space-basedinterceptors under the ABM Treaty. Any interceptor that moves is prohib-ited. So is a space-based interceptor. An auxiliary point, which is nottreaty, is how to test missile defense systems. It was very difficult to figureout a treaty-compliant way to test SBI ten years ago. But it is not justspace-based interceptors, it is any satellite used for warning or to track ordiscrimination or anything like that. We never actually got an agreementwith the Russians as to whether the DSP satellites could be used, whetherthey are treaty-compliant or not. We defined satellites as being adjunct tothis or that radar system. That was good enough for development, butthere was never any agreement on deployment. There are important sys-tems that I have not had the time to talk about here today, such as navalsystems, navy theater wide, etc. which could not be made part of an inte-grated missile defense system. You certainly couldn’t test it; it was not evenpossible to test an Aegis cruiser as part of a recent test. What you findwhen you look into it is that the ABM Treaty permeated every piece ofthinking and people were self-deterred in development and in testing. It isgoing to take a while to get over that, to just try to go and work the prob-lem with the best technologies rather than the ones that we thought theRussians wouldn’t object to or that Congress would modify for treaty pur-poses or whatever. When one can forget those previous treaty constraintsand just attack the problem, it opens up an enormous number of options.

23

Q: I am Baker Spring with the Heritage Foundation and I will be speakingon that topic, chapter and verse, in next month’s Roundtable. I think theoverarching point to make here is that the treaty had as much of a nega-tive impact on the way we looked at the technology as it did in the individ-ual restrictions that it imposed on it directly. That is, it created a culture inwhich there is no light at the end of the tunnel on any particular techno-logical avenue, that you’re going to find, that they’ll put in the contract withNTW within DOD, that the motivation to do it right and do it first and getit done, was nearly undermined. So in my opinion, the ABM Treaty, as itis related to restrictions on testing and development, was “arms control byfrontal lobotomy.” All we would do is just pretend that progress isn’tpossible and then make it a self-fulfilling prophecy.

1 R.Walpole, “The Ballistic Missile Threat to the United States,” NIO for Strategic andNuclear Programs Statement for the Record to the Senate Subcommittee on InternationalSecurity, Proliferation, and Federal Services, 9 February 2000.

2 G. Canavan, Missile Defense for the 21st Century (Heritage Foundation, Washington,DC: March 2002).

3 G. Canavan and E. Teller, “Strategic Defence for the 1990s,” Nature 344 (19 April,1990): 699-704.

4 D. Baucomb, MDA Historian, “The Rise and Fall of Brilliant Pebbles,” InternationalFirst Flight Symposium, 2001.

5 S. Nozette, “Clementine Mission,” Science 226 (16 December 1994): 1835-18626 G. Canavan, “Space-Based Missile Defense and Stability,” Proc. American Philosophical

Society Annual General Meeting, 29 April 2000.

24

RECENT WASHINGTON ROUNDTABLESON SCIENCE AND PUBLIC POLICY

Henry Cooper — Defending American from Offshore Missile Attack (October, 2002)

Margo Thorning — Climate Change and Public Policy: European and American Approaches (September 2002)

Baker Spring — Missile Defense in a World Without the ABM Treaty (June 2002)

Gregory Canavan — Update on Missile Defense Technology(May 2002)

Dorothy Denning — Is Cyber Terrorism Coming? (May 2002)

Jesse Ausubel — Does Energy Policy Matter? (April 2002)

Bruce Ames — The Causes and Prevention of Cancer: Do Federal Regulations Help? (March 2002)

James Schlesinger — Responding to National Security Threats in the Post 9/11 World (February 2002)

Bernard Cohen — Risk Analysis: Radiation, Reactor Accidents and Radioactive Waste (November 2001)

Robert Pollack — Stem Cell Research, Cloning and Gene Therapy: Medical Ethics in a DNA World (October 2001)

The Marshall Institute — Science for Better Public Policy

Board of Directors

Robert Jastrow, ChairmanMount Wilson Institute

Frederick Seitz, Chairman EmeritusRockefeller University

William O’Keefe, PresidentSolutions Consulting, Inc.

Bruce N. AmesUniversity of California, Berkeley

Sallie BaliunasMount Wilson Institute

Thomas L. Clancy, Jr.Author

Will HapperPrinceton University

Willis HawkinsLockheed Martin (ret.)

Bernadine HealyCleveland Clinic Foundation

Charles KrauthammerSyndicated Columnist

John H. MooreGrove City College

Robert L. SproullUniversity of Rochester (ret.)

Chauncey StarrElectric Power Research Institute