Missile Defense in Modern War

Gregory H. Canavan

It is now agreed that the US could be subject to missile threats within a decade; the Rumsfeld Commission argues that attacks from "rogue" states could occur in a few years. As the timing of these developments is uncertain, and the credibility of the organizations responsible for monitoring them has been eroded by recent events, a prudent hedge appears to be to provide some capability as soon as possible for as wide a range of threats as possible. However, it is also prudent to consider how those threats could evolve in time to assure that these development paths are adequate for a range of possible attacks. While nuclear and other weapons of mass destruction are likely, terrorist attacks from within, chemical and biological weapons, short-range missiles from ships close to shore, and others are no less pressing.

Missile threats will grow as long as there is no effective counter to the free ride they now enjoy. Planned defenses rely on the search and interceptor technologies found to be inadequate for similar threats in earlier decades. Satellites can provide better detection and warning, but cannot compensate for inadequate basing. Space- or ground-based boost-phase interceptors could overcome these problems and apply pressure to all phases of attack, but are not in development. Versions of each based on current technology would not significantly impact strategic systems. Their joint development could promote cooperation on a range of missile defense issues.

Theater missile defense developed in parallel with strategic, aircraft, and cruise missile defenses. While theater systems do not have to conform to ABM Treaty limits, their demarcation is not clearly defined, so theater interceptors were largely stripped of their ability to intercept missiles. Even so, the Patriot's residual capability had some effectiveness against missiles in the Gulf War and maintained Alliance morale and integrity. Patriots were directed by C-band phased array radars of a few meter aperture, tens of kilometers from the intercept. Even with semi-active "track via missile", that only gave resolutions of meters, which meant Patriots had to carry large warheads and fuse them precisely to closest passage. That made them sensitive to timing errors and unintentional evasive maneuvers executed by SCUDs breaking up during reentry. Those problems have been addressed by PAC 2's improved fuse and PAC 3's on-board K-band radar, which directs intercepts and provides precise angle and range measurements for higher order guidance for maneuvering targets.

Theater High Altitude Air Defense (THAAD) and Navy Theater-Wide (NTW) interceptors are designed to extend defended footprints by intercepting missiles outside the atmosphere, which makes lower-leakage multi-layer defenses possible. Both are commanded by on-board infrared cameras, which reduces guidance package mass and provides the higher bandwith needed for hit-to-kill intercepts of high velocity missiles-at the price of range measurements, which are not possible with passive sensors. On-board guidance also frees THAAD and NTW's large, high precision X-band radars from the need to command intercept end games, so they can search for long range targets and generate precision metrics for discrimination. Measurements of radar length, area, and motion-combined with infrared measurements of physical size and temperature, should largely eliminate the effectiveness of decoys in theaters. A 600 km SCUD has a maximum altitude of ~150 km, where the radar could detect light decoys by their differential deceleration by the atmosphere. Thus, only decoys that could replicate all radar dimensions of weapons would be effective. Such decoys are complicated, heavy, and reduce payload significantly, so aggressors could not be assured of their effectiveness.

THAAD is still in development. It has had more than the usual number of problems, which have been well publicized. It was unsuccessful in six attempted intercepts; however, the failures have been caused by faulty thrusters, connections, and other quality control problems. None suggest a fundamental flaw in its interceptor, sensor, and guidance, which appear adequate for the conventional, non-maneuvering targets for which they were designed. The combination of infrared guidance and hit to kill has had significant testing. The Homing Overlay Experiment (HOE 1984) used it to intercept a reentry vehicle at intercontinental range and speed. The FLAGE (1987), ERIS (1991), and ERINT (1993 and February and June 1994) intercepted tactical missiles. The US-Israeli Arrow interceptor, which complements an infrared focal plane with a ground-based radar range measurement, has achieved several hits, at least one direct.

These defenses should be adequate for unitary conventional weapons. Nuclear weapons require much lower leakages. Theoretically, they could be obtained by compounding the kill probabilities of the above concepts, but in practice they would probably require more direct counters. Alternatively, attackers could use multiple submunitions=bomblets containing chemical or biological weapons. It is possible to release such munitions within seconds to minutes after the completion of boost phase, which is far out of range of Patriot, THAAD, and NTW. Early dispensing of munitions could saturate, exhaust, or render cost- ineffective current theater defenses. At theater speeds, reentry protection requires little more than a coat of plastic on each cannister, so there are few technical barriers to dispersal after boost, which actually benefits explosive, chemical, and biological weapons, whose effectiveness increases with the total area covered and hence with the number of munitions.

Theater boost phase.The general solution to bomblets is to push intercepts back up the weapon's trajectory: to pre-launch or boost phase. However, success to date in locating theater missiles prior to launch is not encouraging. There are concepts for intercepting theater missiles in boost: air-or space-based lasers, ground- or space-based kinetic interceptors, and less-developed concepts. Each has its own technical and cost issues. Suppression of launches from a 1,000 km country could require 3-4 ABL (Air Based Laser) orbits. It takes 3-4 aircraft to keep one on orbit continuously, so for costs of ~$300M per ABL, inveestment cost would be ~$3-5B per country.

Space-based lasers (SBL) are well suited to counter fast burning theater missiles. As they avoid most of the atmosphere, propagation is less of a concern. However, target hardening remains an issue, and the vulnerability of the SBL itself becomes one. Its location, capability, and hardness are subject to inspection and test, so it might be possible for a rogue to attack the SBL with a lower level of technology. Because satellites move continuously on orbit, at any time most are somewhere else; thus, just the fraction overhead must be able to handle the number of missiles launched. Those "absent" cannot participate, which is generally viewed as a penalty, although if the SBLs elsewhere are over other areas facing missile threats, they would contribute to global defenses. About 20 five megawatt hydrogen fluoride lasers with 4 meter mirrors ("5-4 SBLs") would be required to cover the 30 degree latitude "SCUD belt" against the launch of a missile burning for 60 sec and hardened to 10 KJ/cm2. Theater launches are point like, so the number of SBL scales directly with the product of number of missiles and their hardness and inversely with their boost time and SBL brightness. Deploying technology in space is more difficult than on airplanes or ground platforms, and SBL development has been sporadic, so space lasers are about a decade less mature technically than airborne lasers. SBL costs scale roughly on brightness, and those in the 5-4 class might cost ~$0.5B each, so this constellation might cost ~$10B, which is much smaller than the costs estimated for strategic constellations, primarily due to the much lower launch rates.

Space-based interceptors (SBI) are small rockets, launched from space platforms, with sensors for self-directed hit to kill. A missile requiring 90 sec to accelerate could be reached by a promptly launched SBI with an acceleration of 10g's and a top speed of 10 km/s from a distance of r ~ 400 km. Each SBI platform could cover an area ~ pr2 ~ 500,000 km2, so about 2pRe2/pr2 ~ 500 such SBIs could cover the SCUD belt. Ranges are monotonic in acceleration: 5g SBIs could reach 90 sec missiles from about 200 km; 20g SBIs from about 800 km. The optimal combination of acceleration and velocity can be determined by trading off the greater number of slow SBI platforms that would be needed against the greater mass of faster SBIs. A 5 kg kill vehicle costing $5M with launch costs of $10K/kg produces a ~470 platform constellation of 10g, 7 km/s SBIs costing about $4B. If more than one missile is launched in a short interval, constellation size and cost increase proportionally. Iraq launched several SCUDS simultaneously in the Gulf War. That would be more difficult with longer range missiles. For longer range missiles SBI range and coverage increase roughly in proportion to missile range. Thus, absenteeism drops rapidly, making SBIs attractive and intra-theater attacks against allies unattractive.

Fast ground-launched GBIs with high accelerations and velocities can reach missiles in the boost phase in some theaters. It has been noted that "short- burning Scuds could be destroyed by small homing interceptors.from as much as 50 km distance from the launch site." With higher acceleration interceptors and longer burn times, larger ranges are possible. In Asia, engagements involve launches over large bodies of water, which could allow interceptors to be placed close to the missile launch area. If a "fast" 7g, 7 km/s GBI was placed between the launch area and the target and fired promptly on missile launch, the GBI could reach the missile by burnout from a range of ~760 km. The cost of a deployment per country, which would be dominated by launch platform, command and control, and operation during periods of crisis, could be ~$1B.

Unmanned aerospace vehicles (UAV) or pilotless aircraft carrying kinetic energy interceptors could in some geometries reach missiles in boost. If they remained behind front lines, their timelines would be similar to those for GBIs. However, it would be difficult to carry a fast interceptor on a UAV, so from the rear, they could only reach forward-deployed missiles. If dispersed over the missiles before launch, they could intercept a wide range of missiles in boost even with modest interceptors. However, that requires the violation of the opponent's airspace before operations, which is provocative and assumes a degree of survivability that is unlikely with either current UAVs or affordable future ones. The cost of the UAV and interceptor would be ~$1M. Overall costs would be dominated by command, control, and operations.

Assuming successful development, the concepts above can be roughly compared on the basis of their availability, coverage, robustness, and cost. Due to extensive prior development, SBI and GBIs could probably be available in 3-5 years, which is about when the inter-theater missiles for which they are best suited should appear in numbers. Their lethality is robust, and they should be difficult to decoy. UAV overflight is a cheap, near-term solution with significant political costs and risks due to platform vulnerability. SBIs' greatest sensitivity is to near-simultaneous launches and short burn missiles, which increase constellation number and cost nonlinearly. The cost of GBI and SBI defenses per country against inter-theater missiles are ~ $1B and $5B, respectively; thus, GBIs would be preferred to SBIs for trajectories they could access. For ~1,000 km missile ranges, SBIs would be cost competitive with ABL, if given credit for global coverage. For ranges over 2,000 km, SBIs would be preferred even if they were not given credit.

Battle management. Recent improvements in the internetting of sensors, battle management, and command and control have significantly multiplied the effectiveness of the interceptors described above. As little as a decade ago, it was not clear that the flood of sensor data from satellites, Patriot, Aegis, and THAAD radars could be combined computably. That is now done routinely in real time on operational platforms. The Gulf War was a watershed in information fusion as much as in combat tactics. It is now possible to internet and fuse as many radars and satellites as necessary to synthesize defended footprints as large as the theaters our troops operate in and the continents our allies stage from. An essential element of that progress is the development of the corps of highly skilled men and women who are the heart of these defenses.

National Missile Defense (NMD). The end of the Cold War and the transformation of the USSR into Russia shifted the threat from all-out strategic launches to accidental or unauthorized launches plus contingencies. Third world developments such as the North Korean missiles, China-Taiwan straits, and India-Pakistan tests have since complicated that picture. It is now generally agreed that within a decade the US could be subject to global missile threats with nuclear and other weapons of mass destruction with competent radar and optical penetration aids. The Rumsfeld Commission argues the timelines for those threats could be as little as a few years.

Ground based systems. That the Earth is round is a fundamental limitation on ground based systems, because in intercontinental attacks, the weapons are hidden from ground-based radars until they come within a few thousand kilometers. A radar in the center of the US can only see slightly beyond each coast, so interceptors committed by the ABM Treaty permitted configuration could only protect a fraction of its interior and none of its coast. Distributing the interceptors alone would not alleviate this sensor limitation. Ballistic missile early warning system (BMEWS) radars at the perimeter of the US can see objects a few thousand kilometers from the coasts, which is about the distance an interceptor would have to fly out from the center of the US to intercept them at the coast. However, the incoming weapon is at intercontinental speed, while the interceptor's average speed is about half that, so the weapon would arrive first. Central basing of GBIs would not protect the contiguous, let alone the non-contiguous states.

Radar performance degrades in environments disturbed by nuclear explosions. Hit- to-kill GBIs eliminate the nuclear weapon in the interceptor, but not that in the incoming RV, which could detonate on contact or command. That would produce widespread ionospheric disturbances that could interrupt radar or infrared sensors for times longer than the attack. The US has no relevant data on nuclear phenomenology at relevant intercept altitudes. While x-band radars are less susceptible to nuclear blackout, the Achilles heel of Sentinel and Safeguard was random refraction from multiple bursts, for which there is no experimental evidence. For attacks greater than a few weapons, this introduces a fundamental uncertainty into NMD.

NMD GBIs use solid rockets, passive infrared focal plane guidance, and hit to kill similar to those in THAAD and NTW. Their larger rockets produce velocities approaching the RVs to defend footprints of several thousand kilometers. They intercept exoatmospherically, where RVs are still cold, hence they must use more sensitive focal planes. Even with these extensions, GBIs only take advantage of only the last few thousand kilometers of the descending phase of the the RVs' trajectory. GBIs depend on detection and discrimination by ground based radars and satellites for target and trajectory information, which degrades in an unknown manner.

Non-nuclear weapons of mass destruction are possible. The damage from chemicals would hardly justify intercontinental launch, but biological attacks are so damaging that their use as intercontinental payloads has been described as "a likelihood." Per unit mass, biological agents are as destructive as nuclear weapons. They can survive transit through space and reentry. And they can do so subdivided into a large number of individual cannisters. The first makes them as serious as nuclear weapons. The second makes them a real threat. And the third makes them more difficult to defend against, as they represent a payload fractionation mechanism that could exhaust planned defenses. The cannisters could be too small and cheap to be effective targets, even if GBIs could detect and discriminate them. It is not clear how to address this threat. Smart rocks on GBI are a possibility, but in addition to the economic issues, leakage restrictions for effective defense of population would be difficult to meet. Even nuclear intercepts would be marginal, as lethality against biological agents requires ranges of ~100 m for neutron kill and 1,000 m for thermal and shock, while cannisters are dispersed more widely for efficient attacks on cities.

GBIs committed on the basis of DSP or SBIRS detection and track could intercept weapons about mid-way. That would provide global coverage that was largely independent of interceptor basing, although only with single-phenomenology infrared discrimination in the region where optical decoys are most effective. If satellites serve only as adjuncts to early warning or battle management radars, the required BMEWS confirmation of satellite detection would occur about midcourse, which would give GBIs ~15 minutes to fly out. At an average speed of ~3.5 km/s, that gives a range of ~3,000 km, which would still require the GBIs to be based on both coasts. Thus, satellite commitment and direction of interceptors are essential elements of robust global coverage. Due to the uncertainties in threats, discrimination, and nuclear environments, the prediction of the performance of current NMD concepts is uncertain for attacks larger than a half dozen nuclear weapons and questionable for cannistered biological threats, for which they were not designed. Thus, current NMD concepts would produce defenses with some new technologies but essentially the same weaknesses as those deployed and found inadequate against similar threats three decades ago. As current NMD concepts are represented as the best that can be done with US ground-based decending phase systems, it is appropriate to relax those restrictions to see what other concepts could contribute.

Boost phase defenses largely eliminate decoys and disturbed environments and operate before multiple weapons are deployed, so they could address the major uncertainties. "Pre-boost," i.e., destroying missiles before launch has not been effective in theaters, where launchers are difficult to find, but should be possible for intercontinental launchers, which are larger, fixed and take significant infrastructure and time. For those reasons, preemption is perhaps technically the most feasible solution for rogue threats today, but it involves action before the initiation of hostilities and is strategically destabilizing, so it is not elaborated in the discussion below, which concentrates on boost phase concepts.

Fast GBI (Ground Based Interceptor). A typical liquid-fueled, 3g intercontinental missile has a boost phase of ~270 seconds and burns out at about 490 km altitude 780 km downrange towards its target. An interceptor launched without delay from along its track with an acceleration of 7g and maximum velocity of 7 km/s could reach it by burnout from a distance of 2,300 km, which is larger than the distance to assumed threats such as North Korea and the Middle East. The interceptor flyout distance is ~1554 km, which is a survivable standoff range. It could be command guided, look for a bright plume rather than a dim body, and use laser hard body handoffs to reduce weight, cost, and signature and increase survivability. Such high-acceleration interceptors have been developed: Sprint produced 100g with 40 year old technology. Fast GBIs could be launched on DSP for detection and track, although SBIRS higher frame rate would be preferable. Either could discriminate any decoy short of a full first stage missile. Any delay between missile and GBI launch reduces range. A ~60 s delay for multiple observations and characterization reduces the range of a 7 km/s GBI ~20%, which corresponds to a 35% reduction in the area covered. The reduction in GBI flyout distance is ~30%, which reduces the standoff area available for survivability by 50%. The defense would have a kill probability less than unity; thus, it would not be a single-layer, stand-alone defense. However, it should put enough pressure on the boost phase to complicate attack planning and reduce the threat faced by downstream layers enough for the overall defense to have adequate performance and reliability.

Space based interceptors (SBI) were developed to perform boost phase intercepts over launch areas inaccessible with ground-based systems from survivable, non- provocative platforms. They provide a maximum of autonomy and a minimum of sensitivity to uncertainties in natural and disturbed environments as well as a capacity for intercepts in all phases of the defense. For current threats, survivability and autonomy are no longer as essential, but the ability to execute boost phase engagements and insensitivity to environments are. The requirements and timelines for SBI are much the same as for the fast boost phase GBI, as they require roughly the same acceleration and velocity to reach a given missile before burnout. SBIs' additional degree of freedom is that its platform constellation density can be adjusted for the most efficient combination of SBI velocity and acceleration, whereas the fast GBI's velocity and acceleration are constrained by the standoff distance required for survivabililty.

The command and control, onboard sensors, and response times for NMD SBI are similar to those discussed earlier for TMD SBI. An optimized SBI has an acceleration of about 5g, a speed of 7-8 km/s, and a range of about 1,500 km due to the missiles' long acceleration time. Thus, its absentee ratio drops to about 30 for single launches. The resulting constellation cost is ~$1B, which is less than that of TMD SBI. However, this absentee ratio is still an order of magnitude larger than that for GBIs; thus, which of the two is preferred depends on whether the GBI can reach the missiles and whether the SBIs are given credit for the global coverage they provide.

Lasers react faster than interceptors, but have greater sensitivity to hardening. Airborne lasers have marginal ranges for deep inland launches, but their mobility should make it possible for them to orbit close to the borders of launch areas. If so, for intercontinental launches, from below the missile's path they could irradiate missiles burning out ~500 km above, because in propagating upward the beam would encounter only a thin lens of turbulence, which maximizes range.

By firing downward, space based lasers avoid most of the atmosphere, so they have ranges of thousands of kilometers. Their dominant scaling is that the product of the number of SBLs and their brightness is proportional to the product of the missile launch rate and hardness. A half dozen 5-4 SBL could negate the launch of a single 10 MJ/cm2, 270 s intercontinental missile. That constellation would also place all launch areas at risk. The constellation could cost on the order of $3-5B. The number of lasers would increase with the number of missiles launched simultaneously and their hardness; however, the challenges of achieving simultaneous launch are greater for inter-continental missiles-as are hardening penalties. SBL for NMD are less developed than airborne lasers or fast GBIs, but should benefit from the commonality of technology and common use of constellations for theater and NMD applications. SBL survivability is a concern in that they could be attacked by SCUD-like missiles fired upward in their path. A combination of SBL and GBI could provide overall survivability, quick response for fast missiles, and inexpensive intercepts of hardened intra- and inter-theater missiles.

Cooperation. Interceptor mobility is a common feature of the promising concepts for intercontinental missiles. It is prohibited by the ABM Treaty; however, none of the concepts uses it in a way that would threaten strategic systems. Thus, it would appear inappropriate to apply this criteria to concepts designed for threats that are not viewed as strategic by signatories. As Richard Garwin has argued, "Such a sea-based boost-phase intercept system is not compliant with the 1972 ABM Treaty; but Russia and the three other parties to the Treaty might well agree to a specific exception, especially if this were combined with progress on lower missile levels in Russia and the United States." It would seem appropriate to waive the issue of mobility and compare the concepts on the basis of cost and effectiveness.

A current problem in NMD is how to provide defenses needed for rogue threats without degrading current defensive agreements with Russia, which views the ABM Treaty as binding on both. The US and Russia are apparently far apart on issues of strategic stability in the post-Cold War world and how the understandings embodied in the Treaty should guide NMD programs. However, discussions at the technical level indicate that were it possible for the two nations to jointly develop and control defenses against those rogue threats, that might free them to resolve bilateral strategic issues on a longer timescale with fewer distractions. Such joint defenses have been raised to Presidential levels in both countries.

The boost phase defenses discussed above are potential candidates for cooperation. The fast GBI should have high effectiveness and adequate survivability against rogue threats but none against a Russian attack, so it would present no threat to strategic systems. SBI should have even higher effectiveness, but its constellation of platforms could be inclined so that the SBIs presented no threat to US or Russian strategic systems. And because of the relaxation of the need for autonomy against such threats, the SBIs could be constructed at a lower level of technology, and hence possibly built and controlled jointly. Such defenses might provide solutions to the threats the US and Russia currently express concern about-and only those threats-which could clarify whether the US wants a defense against rogue threats or an entry into a wider defenses, as suggested, and whether Russia wants to guide US NMD towards permitted defenses or influence domestic politics.

Time never seems right for cooperation. A decade ago the common judgment was that Soviet Union wasn't mature enough; now Russia's drift to the right is cited as making technical interactions unsafe. In the few years at the beginning of this decade in which Russia offered cooperation, the US was occupied with its domestic economy. There is always a reason for delaying cooperative efforts. However, it is possible that US pressure for NMD could prevent the Russian Duma from ratifying START II and allow the General Staff to move back to heavy land- based MIRVed missiles-posssibly the most destabilizing option imaginable-and Russian intransigence on NMD could provide the leverage needed for the US to leave the ABM Treaty. These dangers are real. Cooperation on defenses that do not impact legitimate strategic concerns could significantly reduce them.

Summary and conclusions. If rogue missiles gain the ascendancy, it is likely that they will be used against the US and its Allies. If defenses gain the ascendancy, the US and its allies could maintain sanctuaries from which to provide a stabilizing influence. The best way to slow the development of missiles is to take away their effectiveness. Defenses with fundamental limitations will not serve that function; those discussed here could. It is too early to say which would be the best choice technically; that must await the completion of their development programs. Any of them could be an effective deterrent to theater and intercontinental missiles, if deployed soon. However, absent broader and faster development, that appears unlikely.

Gregory H. Canavan

Los Alamos National Laboratory