discriminating it from anything else, fire control, that is, finding precisely where to shoot; and, finally aiming and killing it by hitting it either with some sort of interceptor or with a form of laser energy. Yet since missiles and warheads present different problems in each of the four stages of flight, killing missiles in any of the phases requires equipment sometimes very different from what is required for the other phases. Hence we will describe the general principles of missile defense - what it takes to kill a missile or warhead - phase by phase.
I. Boost Phase
First, a missile destroyed in boost phase is destroyed regardless of where it is headed. Destroying it protects any and all targets it might have attacked. Hence, destroying any missile during boost phase defends everyone. It is an act of global defense. Second, a missile destroyed during boost phase never gets the chance to deploy multiple warheads. Boost phase defense thins the attack most efficiently, and lightens the load on other layers of defense. Moreover, killing in boost phase is the one and only way of defending against a missile that carries so many small warheads as to outnumber any conceivable set of interceptors. Third, killing even a few missiles in boost phase regardless of where they are going makes it disproportionately difficult for an attacker to carry out a militarily rational plan. That is because while the attacker must be sure of hitting certain targets, he must take into account the possibility that his missiles headed for “must” targets will be struck in boost phase. And of course, all missiles killed in early boost phase fall relatively close to the places from which they were launched.
In general, killing missiles in boost phase requires less sophisticated sensors, less complex calculations, and less energy delivered onto the missile than does killing them in any other stage. That is because in boost phase, missiles are obvious, slow, and vulnerable.
Nevertheless, killing missiles in boost phase is difficult because of one requirement: all the equipment for detection, discrimination, fire control and killing must be present within range of the boosting missiles, and just at the time when their engines are burning. Hence the effectiveness of equipment for killing missiles in boost phase depends—like the value of real estate—on location, location, location. And since enemies launch missiles either from their own territory, from others’ territory, or from surprise locations at sea, having the equipment at the right time and place can be a problem.
The problem, obviously, cannot be solved by stationing the missile killing equipment on the enemy’s soil, or by flying that equipment where the enemy can shoot it down. Very rarely, the problem is soluble by putting the equipment on the ground - providing that it is on the soil of a country right next to the one from which the missiles are being launched, close to the point of launch, and directly under the trajectory that the missiles must travel to get to their targets. The problem is also soluble by putting the equipment on ships, but only if the enemy must launch his missiles over the sea and from near a coast, and interceptors launched from the ships can hit them before they get out of range. But mostly, killing missiles in boost phase must mean putting all the requisite equipment, in sufficient quantity, in earth orbits from which it can be present constantly, in relative safety, over all the places from which missiles might be launched.
Boost phase defense can occur in two distinct ways. The first, from the surface of the land or sea, uses radars for detection and guidance. The second, in orbit above the earth, uses optical heat sensors and either laser beams or interceptors.
Let us first consider boost phase defense with radars and interceptors.
Regardless of whether one looks at missiles during boost phase with radar or with heat sensors, missiles are easy to notice and hard to miss. Because of their sheer physical size—up to over 100 feet tall and ten feet in diameter—missiles reflect almost as much radar energy as do airliners. That means that the range at which most modern large search radars can detect missiles during boost phase is limited only by the curvature of the earth. In other words, the moment that a missile perhaps 1,500 miles away rises above about 300 miles and thus above the horizon, it becomes visible. But seeing a rising missile with a radar from so far away is not very useful for killing it. By the time a missile rises above the atmosphere, it is going so fast that no interceptor launched at it from the ground could possibly catch up with it. Besides, by the time a tail chasing interceptor catches up with the missile, boost phase is over or nearly over, and the missile is releasing its warheads. That is why radar detection of boost phase missiles is useful only if it happens very close to the places where the missiles are launched, and if the interceptors are located nearby as well.
Once the radar sees the missile, however, there is no difficulty confirming what it is. Only a real missile reflects radar like a missile and accelerates like a missile. Any decoy would have to be another missile. The computer associated with the radar need only do an elementary calculation of the size of the radar target and its acceleration, and compare it with profiles of known missiles.
Almost any modern well-placed radar can provide information accurate enough to calculate where a missile will be at any given time during its boost phase, and hence to launch an interceptor at that point. Because boost phase missiles move slowly, the radar and its computer need not take many rapid, accurate measurements to calculate the point at which to aim the interceptor. Nor will the missile travel far enough or fast enough to deviate significantly from the predicted course.
Killing boost phase missiles does not require that interceptors possess either sophisticated guidance systems or a particularly lethal punch. It does require that interceptors get to very high speeds very fast. Since boost phase missiles are big and slow, the interceptor can home in on the radar energy that the missile reflects. In boost phase, a sophisticated heat seeking guidance system is not needed. Having one can actually be troublesome, because the guidance would have to be programmed to hit not the hottest point—the exhaust below the missile—but rather somewhere above. That requires accurate knowledge of what various kinds of missiles look like to heat sensors. A simple radar sensor is actually better.
But regardless of the guidance system, a surface-based interceptor that goes after a boost phase missile does not have to make terribly fast calculations or tight turns. And when an interceptor hits the body of a boosting missile, it will pierce it much more easily than it would a warhead. Hence there is no need for the interceptor to carry any warhead of its own or even to be tipped with a dense spearhead material, such as depleted uranium. Once any interceptor or its fragments hit a boosting missile, it will explode.
A. Boost phase defense using heat sensors, space-based laser, and space-based interceptors
In all discussions of sensors or weapons in orbit around the earth, we must keep in mind the basic facts of orbital life: While any given satellite rotates around the earth, the earth is rotating underneath it. The satellite in an orbit 200 miles high goes around the earth in about an hour and a half. During that time, the earth will have rotated one eighteenth of the way around its axis. That means the satellite will not again pass over the spot from which it was launched (or indeed again over any given spot on its course) until eighteen hours later. And that means that if one wishes to have one such satellite in view of every spot on their orbit at any given time, one must keep eighteen of them in orbit at all times. The only kinds of satellites that are over one area of the earth at all times are the ones that orbit the earth 22,300 miles away. But from such distances, very little precision sensing can be done, and no weapons at all can be used.
Although it is technically possible to build radars for work in orbit, in practice radars are too heavy and require too much electricity to be useful in outer space. That is doubly true of the long range, high-resolution radars that are needed for anti-missile work. That is why space based sensors are heat sensors. They work by means of telescopes that focus faint infra-red (heat) energy onto highly sensitive cells connected to computers.
During boost phase, the engines of missiles generate temperatures up to 2,300 degrees centigrade, highly visible light, and incredible amounts of high energy and short-wave infra-red radiation. Heat sensors stationed on satellites at geo-stationary orbits 22,300 miles away can feel increased heat even without having their temperature artificially brought down to near absolute zero. Such un-cooled sensors are relatively simple, cheap and reliable. The effectiveness of such satellites for detection of boost phase missiles depends on three factors. First is how accurately the satellites are calculating their own position over the earth. The satellites’ constant calculation of their own position is necessary so that, when they calculate the position of the boosting missiles relative to themselves, they can then translate that to an absolute location over the surface of the earth, and communicate that to other sensors. Second is the rate at which the satellites’ sensors take images of the boosting missiles. The faster they do it, the more points at which they see the missiles, the more accurate will be their prediction of where the missiles are going and where they will be at any time of their flight. Third is the amount of data-processing that goes on aboard the satellites: how well and quickly the data is transmitted to other sensors that can take action with regard to each missile.
For the purpose of detection alone, however, even the crudest satellites of the 1960s performed these three functions more than sufficiently. The advent of Global Positioning System satellites simplified and increased the accuracy of geo-location. At the same time, increases in computing power have enabled satellites to perform functions formerly reserved to processing centers on the ground.
To adequately discriminate missiles during boost phase, the crudest satellites of the l960s needed upgrades. Essentially they identified missiles by the fact that they were the only sources of energy that were accelerating and gaining altitude rapidly. As knowledge of the thermal characteristics of missile exhausts improved, it was fed into these satellites’ computers. For modern satellites that possess vast data bases and sophisticated formulas for recognition, however, boost phase discrimination is perhaps more of a non-problem even than boost phase detection. The only question is whether these satellites communicate the results of detection and discrimination to some other sensors for fire control first, or send them directly to space-based weapons. That is because—-for boost phase only—-both space-based lasers and interceptors are perfectly able to do their own fire control.
Both space-based lasers and space-based interceptors would possess sufficiently accurate sensors and more than enough computing capacity to calculate the boosting missile’s track accurately enough to aim the interceptor toward a predicted impact point or to aim the laser beam to the target. However, the difference between the capacity of space-based interceptors and space-based lasers for fire control start from the fact that since the interceptors are small, they necessarily carry sensors with a shorter range and a narrower field of view than the sensors carried by the space-based lasers. That means that the satellites at geo-synchronous orbit that detect and discriminate must tell the interceptors rather precisely where to direct themselves, so that the missiles will end up within the range and field of view of their sensors. Hence, to enable space-based interceptors to do their own final fire control, the detection-discrimination satellites must be quite accurate themselves. In addition, since each interceptor can hit only one missile, the detection-discrimination satellites also would have to sort their data on missile targets and decide which individual interceptor or interceptors would have the best chance to hit which missiles. Only then could the satellites decide what information to send to each interceptor. That means that satellites that do fire control for interceptors must be reasonably sophisticated. The alternative is to program each interceptor simply to head for whatever boosting missile happens to come within its field of view and accept the resulting duplication and inefficiency.
Fire control for space-based lasers is a much simpler matter because space-based lasers could see boost phase missiles with the aid of their own telescopes—telescopes that would be almost as sensitive as those of the satellites at geo-synchronous orbit but much closer to the missiles. Hence space-based lasers would have the option of doing their own detection and discrimination. Being able to see boost phase missiles nearly as accurately as the geo-synchronous satellites, but from not nearly so far away, they would be better equipped than any other sensor to calculate the missiles’ track. Since each SBL would know where every other one was, each could calculate for itself which missile it could engage most efficiently and in what sequence. Since several SBLs would be looking at part of a changing mix of boosting missiles, each would have to take into account the presence of the others when making its decision on firing priorities. Thus their calculations would be straightforward.
The amount of energy needed to kill a missile in boost phase is low.
A space-based interceptor could kill a missile in boost phase by colliding with it. Creating collisions between objects in space first became possible in 1960, when the first orbital rendezvous was achieved. As early as 1962, the concept of Ballistic Missile Boost Intercept (BAMBI) was developed. Pods of interceptor rockets would have been stationed in orbit. When offensive missiles were launched within the range of any given pod, that pod would have launched its interceptors. The interceptors would catch up with the missiles and get close enough to kill them by exploding nuclear weapons. Over the years, improvements in the guidance systems of interceptor rockets made it possible to plan on using conventional warheads. By the mid-1980s, small, light, accurate guidance systems made it possible to do away with warheads altogether, and to create actual collisions between interceptor rockets and missiles. By the late 1980s, yet further improvements in guidance, communications, and data processing made possible the ultimate refinement in the concept of space-based interceptors: No longer would the rockets be stationed in pods that orbited along roughly the same path along which missiles would travel, and catch up with the missiles. Now they would rotate around the earth counter to the direction from which the offensive missiles would be coming, and be spread individually along their orbital paths. The interceptor rockets would not have to catch up. Each would just divert itself into a collision course with a missile. At closing speeds of some 12 kilometers per second, the collision would shatter both the interceptor and the missile. Nevertheless, in the 21st century just as in 1960, space-based interceptors would work only if they were in the right place at the right time.
While improvements in technology made these kill vehicles smaller, lighter, and more lethal, they could not change the fundamental feature of space-based interceptors, namely the amount of space that any given interceptor can cover. That is because there is an inflexible limit in the acceleration that any rocket can achieve. The maximum speed of any rocket (over and above any speed at which it might be traveling when its engine is lit) depends on the speed of the gases that exit from its engine. That maximum is set by the speed at which the fuel burns and its gases escape from the rocket’s rear end. It turns out to be around 6 kilometers per second. The practical question then becomes how far can any interceptor rocket deviate from its course to smash into a missile? The answer depends in part on the accuracy and sensitivity of the fire control sensors that can tell each rocket where to direct itself so that its own on board guidance system can create the intercept. In practice, each interceptor rocket might be able to cover a cone with a radius of some 500 kilometers.
To create even the thinnest of screens over the earth would require some 800 space-based interceptors. How many interceptor rockets to place in view of any given point of the earth from which missiles might be launched is a simple matter of geometry and arithmetic.
(2) Space-based lasers
Laser light occurs when various kinds of molecules are energized, either electrically or chemically, so that outer electron shells are almost torn away. As the molecules return to their normal state and the electron shells snap back, they release particular wave lengths of energy. By contrast, natural light is a mixture of wave lengths (colors) and can be broken down into its components. Because laser light is of a single frequency, its waves can be shaped to hit a target with a single, flat front. This makes it possible for lasers to deliver their energy onto well defined surfaces in a penetrating manner. Depending on their power and wave length, lasers can read a supermarket purchase, operate on an eye, or cut metal.
Since missiles are barely strong enough to withstand the heat and acceleration of boost phase, lasers can cause them to self destruct catastrophically by quickly depositing onto their bodies more energy than they can take. The energy of tens of light bulbs per square centimeter for about a second is usually enough.
Laser light, like other light, can be projected (for practical purposes instantaneously) in concentrated doses over long distances. The presence of air, however, diffuses all light. Any and all light travels freely only in outer space. How tightly concentrated the beam is depends on the size of the mirror that projects it. The bigger the mirror, the tighter the focus it can generate at any given distance. But for any mirror of any given size, a beam will be four times as concentrated on a target that is half as far away (or 16 times as concentrated on a target that is only one fourth as far away.)
Although there has been much discussion and experimentation with other kinds of lasers, modern missile killing lasers generate light by burning high energy fuels like hydrogen and fluorine, or oxygen and iodine. Since this requires a rapid drop in pressure after combustion, the process works best in the natural vacuum of outer space.
The development in the 1980s of chemical lasers that develop two to ten million watts of power went hand in hand with the development of technologies for focusing beams with large (four to ten meter) mirrors to distances of several thousand kilometers, and with technologies for pointing those beams accurately at those distances. Thus since the 1980s the elements for missile killing, space-based lasers have existed.
Each SBL could kill a boost phase missile with a discharge from one to a few seconds (depending on the power of the laser, the distance to the missile, and the hardness of the missile.) Each could carry enough fuel for up to a hundred shots. Since each could kill at ranges of a few thousand kilometers, perhaps a dozen at altitudes of 500 kilometers could have at least one over every part of the earth at all times. The practical question would be: how many would we want over which part of the earth at any given time?
(3) Airborne lasers
It is theoretically possible to mount a chemical laser on an aircraft, fly it at high altitudes near launch sites for ballistic missiles, and destroy those missiles as they are launched. But there are the following problems. a) Generating a chemical laser beam requires a vacuum. It is possible, by controlled explosion, to create a vacuum around the combustion chamber on the aircraft; but that takes machinery, creates vibrations, and permits lasing only so long as the vacuum lasts. b) The instabilities and vibrations of aircraft make it difficult to point a laser gun accurately at long distances. c) The atmosphere through which aircraft borne lasers must propagate is an unpredictable medium and imposes restrictions on how far a laser may propagate missile killing energy. d) Since the airborne laser must be relatively near the missile launch site—often over enemy territory—it is liable to be destroyed by enemy surface to air missiles or fighter aircraft. Hence, airborne lasers would be useless against any adversaries who had even minimal control over their air space. e) The notion of flying an aircraft continuously and indefinitely in harm’s way while waiting for something to happen is fantastic.
II Post Boost phase
The small engines of post boost vehicles (PBVs) of missiles do not emit enough energy to be seen by uncooled sensors from geo-stationary orbits 22,300 miles above the earth. These sensors will, however, calculate the position of PBVs from their earlier observation of the trajectory of boost phase missiles, and will pass the information on to other sensors. Hence, detection of PBVs must be a cooperative effort. Since, except for shorter range missiles (or missiles launched from the sea), Post Boost Phase must take place over or near the territory of the country that launched the missiles, the sensors that deal with PBVs must normally be based in earth orbit. Occasionally however a radar that is based close to the origin of the missiles may detect and track PBVs. But only space-based heat sensing devices are capable of discriminating PBVs.
Although the engines of PBVs are much smaller than those of boosters, PBVs are still unmistakable because they have a unique combination of heat emission and movement. Ever since the 1980s, a variety of US sensors has compiled an extensive record of the heat emissions and movements of most of the world’s PBVs. This data base theoretically would make it possible to tell not just that an object was a PBV, but precisely of which kind. Discrimination would also be aided by correlating the thermal signature of the PBV with that of the booster from which it had come.
Although PBVs deviate from strictly ballistic trajectories in order to dispense their warheads and aim them at specific targets, they deviate only within a degree or so. Since an interceptor need only be directed to where its target will be within some ten degrees of where it is looking, the PBVs’ variations in course are not enough to complicate the targeting of interceptors against them. If the PBV is to be attacked with lasers, its variations in course are irrelevant because the beam will hit the PBV instantaneously. Moreover, whether the PBV is to be attacked with lasers or interceptors, the small engines, unlike those of boosters, do not put out enough heat to mask the body. Hence targeting a PBV is actually easier in some ways than targeting a booster.
Whereas collision with an interceptor or its warhead fragments would destroy a PBV catastrophically, a laser would inflict subtler damage. That is because although many PBVs actually have their instruments unprotected, they—unlike boosters—do not have large fuel tanks under thermal and inertial pressure. Lasers, then, cannot simply weaken a PBV until it self destructs. Often, the laser will not hit a fuel tank at all, and must limit itself to destroying the PBV’s instruments. This may prevent the warheads from being armed or dispensed. In any case, a laser kill of a PBV may not be confirmable.
III Midcourse Phase
Since a warhead in midcourse may be going anywhere within a broad swath, to kill it is to protect everything in that swath. The swath defended spreads out from the points where the intercept may be made, so that the broad area behind the possible targets is defended as well as the targets themselves. Hence, though midcourse defense is not global defense, it is defense of broad “footprints.” Moreover, a warhead killed in midcourse is likely to have its chemical or biological contents incinerated during re-entry into the atmosphere, and any radioactive materials scattered. Indeed, any warhead killed during midcourse cannot burden final, terminal defenses. The earlier in the warheads’ trajectory that midcourse defense can work, the more of the above mentioned benefits will result - and the easier the work is likely to be. Time is of the essence. Midcourse defense is worth doing, but difficult.
Destroying warheads during midcourse is by far the most difficult part of missile defense. Killing is complicated by the fact that lasers are relatively useless against tough warheads. Hence interceptors must do the killing. But the hardest parts of midcourse defense are detection and discrimination. Modern technology however has transformed the task.
The earlier and farther away from the target that detection begins the process of midcourse defense, the better the defense is likely to work. Ideally, the defense could follow each warhead from the moment it left the PBV. But how to put the sensor in the right place? And how to make sure it’s good enough?
Detecting warheads in midcourse is difficult, even when one knows where to look for them. The uncooled sensors aboard the geo-stationary satellites, and even the ones on space-based lasers useful for boost phase defense are useless for seeing objects that at best, are at room temperature. Powerful ground based radars can see warheads. But the problem for American and European (but not Russian, see below) missile defense is that such radars could not be located close enough to the source of the warheads and far enough away from the targets. In other words, because of location, the radars could not see the warheads early enough in midcourse. The obvious solution to the problem of seeing warheads early enough and well enough was to develop a network of heat sensing satellites whose numbers would guarantee that several would be in sight of any warheads any time any where, and whose sensitivity would permit the detection (and discrimination) of warheads. This system has largely been developed. Hence, we must regard radars as primarily of historical and secondary importance. Radar is the original means of detecting warheads in midcourse. For radar to reach far enough to see warheads in midcourse, its transmitters must radiate large amounts of power, and its receiving antenna must be large. The basic capacity of a radar is measured by multiplying the average power of its transmitters by the area of the antenna. By the same token, a radar can multiply this “power/aperture product,” and see farther, to the extent that it concentrates its signal on a narrower area of sky. And it can do that to the extent that another sensor has told it where to look.
The first radars made specifically to see warheads far away were the US Ballistic Missile Early Warning (BMEWS) radars at Clear, Alaska; Thule, Greenland; and Fylingdales, UK. In the 1960s, the Soviet Union built similar structures hundreds of meters long that NATO called “Hen House.” These monsters could see warheads some 3,000 kilometers away. Since that time radar designers have packed many more and more powerful transmitters into antennas. As a result, modern detection radars measure in the tens rather than in the hundreds of meters. The US radars (not the Soviet ones) served only for warning, rather than detection, because they did not transfer the information they gathered to any device that performed the other functions of missile defense, namely discrimination, fire control, and killing.
Detecting warheads by sensing their very faint heat against the cold background of space from hundreds of miles away requires very sensitive instruments cooled to near absolute zero. While the temperature of warheads that have been in space for a short period of time (i.e. warheads that are traveling relatively short distances) typically does not fall far below freezing and thus emits heat in the mid infra-red range (about 5 microns), the warheads of longer range missiles emit weaker, long wave infra-red waves (over 10 microns.) These different targets require different sensors. For detection, the sensors need only distinguish between different points from which the weak heat comes.
The anti-aircraft radars used in the 1950s to discriminate between missile bodies and warheads measured speed. They simply noticed that some objects (the warheads) were not slowing down as they re-entered the atmosphere as other objects were (the spent bodies of missiles and debris.) But, of course, this discrimination was, and is, possible only during terminal phase, not midcourse.
Modern radars discriminate the objects they see in midcourse by measuring the distance to each and every part of each object, and then building a picture. To measure distances accurately, a radar must emit pulses of the highest possible frequency, thereby producing wider bandwidths. Thus radars that emit pulses in the X band (10 million cycles per second), measure ranges with an accuracy of 1.5 centimeters and distinguish between objects as close as 15 centimeters. Hence computers that use information from radars in that class have no difficulty noting the shapes of objects released by PBVs, observing their movements, and comparing the data with what is known of the characteristics of warheads and decoys. The largest X band radars can do this at ranges of some 3,000 kilometers.
X band radars, however, like all other radars, cannot see over the “radar horizon,” the curvature of the earth. That means that the range at which they actually see and discriminate the warheads—how far away they see them from the targets they are supposed to be defending—depends on where they are located and what the range of the missiles is. In most cases involving all but shorter range missiles, the fact that the radar cannot be located on any except friendly territory will prevent it from seeing the warheads until after they have begun their descent onto friendly territory.
Heat sensors, by contrast, are based in earth orbit. They discriminate objects by measuring the differences in the infra-red waves they emanate. To do this, these sensors do not have to build an image of the object. They only have to be able to register a spectrum of frequencies coming from each, constituting a “signature,” and compare it with a known database. The range at which a sensor with a telescope about one meter wide can receive enough infra-red waves onto its instruments to discriminate is about 1500 kilometers. But the key difference is that these sensors in earth orbit are not limited either by the politics of the countries they fly over, nor—for practical purposes—by the curvature of the earth. That means that such sensors can see and discriminate warheads and decoys as soon as they leave the PBV, before they reach the highest point in their trajectory and start down.
The purpose of decoys, of course, is to fool sensors. Designers of military space objects, as well as of aircraft, have much experience in creating false targets for radars as well as for heat sensors. There has been much discussion of how the best radars may be foiled by enclosing warheads in tin foil balloons, and proliferating empty balloons in space. There has also been discussion of the possibility of deep freezing warheads to bring them to the temperatures of deep space, like decoys. Because these techniques are easier said than done, none is part of any nation’s arsenal. Success in spoofing comes from knowledge of the enemy’s data base and of the enemy’s criteria for interpreting it. Absent espionage, however, this knowledge is usually lacking, and designers of countermeasures succeed only when they happen to guess correctly. The principles of sensors for discrimination, and of the countermeasures to fool them, are known to all. Success by either side, however, is not a matter of principle but of applied technology—and of lucky coincidences.
Since discrimination in midcourse—by whatever means—requires close observation, the sensors that discriminate must also generate very accurate information about the trajectory of each warhead. To generate that information, the sensors must be tied to computers with software written to record and calculate many data points quickly. In addition to such tracking software, the essence of fire control in midcourse is to transmit to interceptors information accurate enough so that they may get within range of their on board sensors in good time.
Indeed, the practical questions concern just which sensors would perform which functions. The answer to this question is so important to midcourse defense because it strongly affects how soon after detection the interceptor can be launched. That in turn determines how far away from the target it will have the chance to make the intercept. The greater the distance, as we have seen, the greater the benefits of midcourse defense. Moreover, an interceptor that is launched early has more time to stabilize itself on the right course, and may carry simpler sensors. Fundamentally, the choice is between doing fire control from the ground, or from earth orbit.
a) From the ground
Before the invention of space-based heat sensors, the only option was for the computers located alongside ground-based radars and interceptors to discriminate warheads and decoys, sort targets, assign interceptors, and guide them either to the intercept itself, or to the point from which they could follow the radar returns from the target, or to the point where the interceptor could follow its own heat sensor to the target. As these computers did so, they made what use they could of the data from the radars and other sensors that had told them that the missiles and warheads were coming. When space-based heat sensors were invented, the natural tendency was to send their data to these computers, so that they could speed up the process of discriminating, sorting and assigning. And in fact it became clear that, thus aided, ground-based computers could launch interceptors toward specific warheads even before the warheads came into view of the nearby radars. Again: the earlier the launch, the better.
But how could engineers make sure that interceptors launched early, and therefore necessarily imprecisely, would find their targets in their field of view, and that—if a decoy happened by—interceptors would make the final discrimination? U.S. engineers, committed by the 1972-2002 U.S.-Soviet ABM Treaty, as well as by inertia, to ground-based fire control addressed this problem by designing guidance systems for the interceptors’ “kill vehicles” that have wide fields of view and an onboard capacity to do limited discrimination. Naturally, it is difficult to pack such capabilities into rockets that have to be small and light. To perform these tasks better than marginally, the “kill vehicles” would require telescopes and sensor arrays bigger and heavier than they can possibly carry.
Such telescopes and sensor arrays, however, are precisely the heart of the space-based heat sensors that make possible fire control from space. Commitment to ground-based fire control leads engineers to plan as if these capabilities did not exist.
b) From space
As we have seen, although radars can discriminate warheads and calculate their trajectories—in their own way and in their own time arguably as well as heat sensors—they cannot do it as early as devices based in space. The dramatic advances in computer power of the 1980s and 90s, made it possible for these complexes of telescopes and sensors to also carry the capacity to sort targets, assign interceptors, order their launch and, after they are launched, to feed each one a stream of data refining its aim point and further specifying the target.
An interceptor thus guided can make do with a small telescope with a narrow field of view and with a sensor that does not have to discriminate. In sum, fire control from space can function entirely without inputs from ground-based radars. In cases where an X band radar happens to be located well enough to see warheads and decoys in midcourse early enough, it can contribute its data to space-based fire control. In principle, it is always better to base fire control decisions on data from more than one source. In most cases, space based fire control would work from only heat sensors. In exchange however, it would gain the great prize of midcourse defense—time.
In order of importance the issues are: getting the interceptor near the target, homing in on it, and destroying the warhead. Let us consider them in reverse order.
a) Destroying the warhead
Before the 1980s, when improvements in guidance made other methods feasible, destroying warheads in mid course could be done only with nuclear weapons. When Soviet dictator Nikita Khrushchev boasted in 1962 that his anti-missile technicians had managed to “hit a fly in space,” he was actually referring to the fact that they had managed to get an interceptor close enough to a target warhead in late midcourse to destroy it with a conventional explosive. Nevertheless, Khrushchev’s technicians did not plan to take such chances against actual attacking missiles. Rather, they planned to get their interceptors within about three kilometers so that the detonation of the interceptor’s nuclear warhead would disable the target warhead.
When a nuclear weapon detonates anywhere, it releases most of its energy in the form of short X rays. When the weapon goes off within the atmosphere, the air immediately absorbs the X rays. The result is the familiar fireball and blast. In outer space, however, the X rays would simply radiate out. If the detonation is close enough to the target, the X rays will simply burn it up, or at least burn up its electronics. Until recently, midcourse (and terminal) kills relied on nuclear weapons not just because interceptors were inaccurate, but because warheads really are in fact hard to kill.
Since the 1980s, increasing accuracy has made it possible for interceptors, reliably, to get close enough for a conventional explosion to destroy the target warhead with shrapnel, or simply by colliding with it. There are two problems, however.
First, since warheads are physically tough, it is by no means sure that whatever hits them, even at speeds of some ten kilometers per second, will destroy them. The material that hits the warhead must be more dense than the warhead, less brittle, and must have enough mass to penetrate. This means that alloys of depleted uranium are the materials of choice. But since their weight limits the range of the interceptor, designers are pressed to do without warheads and try for direct collisions between interceptors and warheads.
Second, though modern interceptor guidance systems are surely accurate enough to get close enough for a shrapnel warhead to do its work, it is by no means sure that they can eliminate all the margins of error necessary to create collisions. That is why, for example, the Israeli Arrow anti-missile system—the only anti-missile system outside Russia—does not rely on collisions but rather uses a shrapnel warhead that is supposed to kill within 50 meters.
The uncertainties of killing warheads lead some anti-missile engineers to long for the certainty that only “non-depleted uranium” can give.
The earliest and crudest way of guiding an interceptor is “radio command guidance.” The fire control system’s radar simply tracks both the interceptor and the target. A computer (or even a human operator) controls the interceptor’s steering mechanism by radio, so that the interceptor’s radar blob is made to merge with the target warhead’s radar blob. Command guidance can be accurate, but it depends on positive radar contact throughout the intercept. And it is difficult to control multiple interceptors against multiple targets.
Modern radar-based guidance works as follows. First, a ground-based radar can bounce its waves off the warheads. The interceptor can follow the returning waves (called semi-active homing) to home in on the warhead. Second, the interceptor could carry its own small radar set to irradiate the warhead and home in on it. However, a radar set that could do the job adequately would be quite heavy. That is because when an interceptor’s sensor looks for the warhead it is supposed to hit, it is likely to see other objects as well. The picture it sees must be accurate enough for it to be able to compare it with the information it has already received from the discrimination sensor, and pick out the target. That is a tall order for the size radar that an interceptor can carry.
Moreover, as we have mentioned, radars are unlikely to be located far enough toward the source of the warheads to be able to see them early enough in midcourse.
For all these reasons, heat sensing instruments are the most efficient means of guiding interceptors to midcourse kills.
Modern heat sensing instruments can create accurate maps of the objects they see, and their computers can compare them to whatever maps they receive from the discrimination sensors, while using tiny amounts of energy. Moreover, space, the environment of midcourse, makes it easy for heat sensing instruments to work. Because the sensors do not have to be exposed and begin looking until the interceptor has entered the cold environment of space, it is relatively easy to supercool them to make them sensitive. Moreover, since midcourse intercept takes minutes rather than seconds and the interceptor need not make sharp moves, the instruments have time to line up with the target.
c) Getting to the target
Assume that a midcourse defense has interceptors well guided and supplied with the information they need to head for and hit their targets. To get to the incoming warheads in good time, while they are still far away and in midcourse, those interceptors must be—in increasing order of importance—fast, launched early, and located as close to the source of the warheads as possible.
The interceptor’s speed is the least important means of making sure that the intercept occurs far away. The slowest interceptor rockets rapidly reach speeds of over 2 kilometers per second. The fastest reach a bit over double that. Since the missile warheads against which the interceptors are launched travel at anywhere from two to seven kilometers per second, they leave precious little time before the intercept must occur. If the interceptor is launched only after the warheads come into view of the radar, the time may be measured in seconds. Hence, an interceptor that is two kilometers per second faster will intercept some hundred kilometers farther away than a slow one. That is a considerable advantage. Nevertheless, it is not enough to reach deep into midcourse for most missiles.
If, however, the warheads are detected and discriminated, and if the interceptor is launched, early, a quantity of minutes is eliminated during which the interceptor is not moving at all, and another quantity of time is added during which it is traveling. Hence, early launch allows interceptors to reach out to ranges that they would otherwise reach only by traveling many times as fast. As we have seen, early launch is crucial to midcourse defense.
The only factor that is more important than going fast and starting early is being nearly there already. That is why the location of interceptors is so important. The least advantageous place from which to launch midcourse interceptors is from near the places being defended. Launching them from there reduces midcourse defense to a kind of long-range defense of a single point. But in fact that point—and lots of others as well—can be defended even better if the defense makes the effort to destroy the warheads as far away from the point being defended as possible. And if the defense decides to do that, it will try to base its interceptors as close to the source of the warheads as possible. A generation ago, before the advent of massive, instantaneous, broadband data transfer, the need to coordinate the work of radars and interceptors made it a practical necessity to locate interceptors next to radars, and to “site” them near the places being defended. Habit (and the ABM Treaty) carried into our own time the notion that interceptors must be based in anti-missile sites. In fact, it is most advantageous to spread the interceptors out to the many places from which they can reach deepest into the attacking warheads’ midcourse.
IV Terminal Phase
Intercepting warheads as they are re-entering the atmosphere above their targets is, like a parachute or a lifeboat, the last line of defense. At best, terminal defenses will intercept warheads some 40 kilometers away, and at worst only some 5 kilometers away. If the warheads are nuclear, and the area attacked is a city, 20 kilometers is the narrowest of margins—no one can have a picnic under such an umbrella—and 5 kilometers merely tempers the disaster. Such margins are barely tolerable to a dug-in military force. But they are still better than nothing. The passengers in the water around the sinking Titanic yearned for cold lifeboats, and those trapped atop the Twin Towers on September 11 would probably have given a lot to take their chances with parachutes.
Terminal defense has all the technical disadvantages in the world, but two advantages: The defense has only the task of meeting objects that are coming directly toward it, and the earth’s atmosphere strips away any decoys. Moreover, terminal defense has a stark simplicity because it can make so little use of any sensors except radar.
By the time warheads reach terminal phase, heat sensing satellites at geo-stationary orbits will have observed the boost phase of the missiles and estimated the warheads’ destination. Long-range radars will have detected the warheads themselves and estimated their destination rather precisely. Perhaps midcourse defenses will have discriminated and thinned out the warheads, and they may even have communicated the remaining warheads’ path to whatever terminal defenses might exist.
Nevertheless, with or without help, the terminal defense radar must see the incoming warheads. By the time that warheads encounter significant air resistance at altitudes of some 40 kilometers and thus begin terminal phase, the warheads coming from intercontinental distances (and thus at shallow angles) are perhaps 150 kilometers away, while the ones coming from relatively nearby will be coming in more steeply and thus be about half that distance. Well beyond such short ranges, even the smallest, most sharply pointed warheads are obvious to modern radars.
But, of course, since the warheads are traveling anywhere from two to seven kilometers per second, the defense has only a few seconds to act.
Discrimination and Fire Control
Although the terminal radar will have detected the incoming objects before they encounter the atmosphere, and may well have information from other systems about which are warheads and which decoys, it will have to observe all of them inside the atmosphere for the few seconds required to measure how it affects their speed. That is the definitive discrimination. At the same time that the computer discriminates, it must generate accurate data on the warhead’s trajectory. The calculations required in terminal phase have to be even faster than those required for midcourse because the warheads’ motion relative to the radar that is observing it is even faster. By the time these calculations are done and the data is communicated to the preassigned interceptors, there will be only perhaps fifteen seconds between the time the interceptor is launched and the warhead hits.
Guidance and Killing
How far away from its point of departure will the interceptor hit the warhead? In terminal phase, the interceptor’s speed, and above all how fast it gets to top speed, are key. When the interceptor’s time of flight is under ten seconds, the fastest ones may keep the warhead as much as 20 kilometers farther away than the slowest ones. That makes a difference, since the warheads are likely to be fused to explode when hit (salvage fusing). When the warheads are nuclear, chemical, or biological, the difference is big.
The interceptors home in on the stream of radar energy that the warheads reflect. They cannot try to home in on the intense heat that the warheads’ friction with the atmosphere generates, because their own high speed through the air heats their own nose cone just as intensely. As engineers have sought to make terminal intercepts at higher and higher altitudes, where air friction is not so overwhelming, they have developed means of minimizing the air friction on the outside of nose cones while super-cooling the insides long enough to allow heat sensing devices to work. But such guidance systems are properly labeled trans-atmospheric, meant for the outer edges of the atmosphere, and therefore for the transition between late midcourse and terminal phase. The usefulness of such interceptors depends on the accuracy of the earliest possible measurement of the different effects of the atmosphere on warheads and decoys. As measurements become more accurate, trans-atmospheric interceptors will become more useful and terminal phase will extend outward—a bit.
Even though terminal phase interceptors do not confront decoys, they have to deal with two peculiar difficulties. First, even if the warhead comes in precisely as predicted, the fact that the interceptor’s guidance system may take a second or so to lock on to its target means that when it does, it will have to make course corrections—sometimes substantial—at very high speed. It will fish tail. That will subject the interceptor to high inertial forces several times as strong as gravity (G forces). If the warhead is programmed to make a corkscrew maneuver during reentry to try evading interceptors even at the cost of losing accuracy, the interceptor will have to make several high speed course changes. Modern interceptors make such changes by firing small thrusters on their lower sides, akin to shotgun shells.
Because the time is so short and the possibility of near misses is ever present, Israeli and Russian (but not US) terminal interceptors do not rely on colliding with the warhead, but rather carry “kill enhancement devices.” The interceptor monitors the shrinking distance between itself and the incoming warhead. The instant that distance begins to widen again, it fires its own device. Timing is crucial. At closing speeds of ten kilometers per second, each thousandth of a second of delay means increasing the miss distance by ten meters. In terminal even more than in midcourse, there is a trade off between trust in the latest devices for measuring time, and the certainties of “non-depleted uranium.”