Shield of Dreams

A critical look at the science and technology required to build an antiballistic system that would make the United States invulnerable to a missile attack

By Tim Folger and Tony Law
Nov 1, 2001 6:00 AMNov 12, 2019 5:39 AM


Sign up for our email newsletter for the latest science news

The handful of people who would be the first to detect a nuclear missile attack against the United States work just outside the town of Colorado Springs. Their daily commute takes them down a sloping mile-long tunnel, past baffled steel blast doors that are 20 feet high, three feet thick, and weigh 25 tons each, into the heart of Cheyenne Mountain. There, surrounded on all sides by at least 2,000 feet of granite, they spend eight-hour shifts in the missile-warning center, one of 15 subterranean buildings arrayed along a three-dimensional tic-tac-toe grid of intersecting tunnels. This is the central coordination facility for the NORAD (North American Aerospace Defense) command.

Data from satellites and a global network of radar stations flow into computers at the missile-warning center, where eight or nine people during a typical shift sit in front of 17-inch monitors. Hardly a day passes without the detection of a launch somewhere around the world, such as the Russians firing SCUD missiles at Chechens or the French lofting a satellite into orbit. But in the 36-year history of Cheyenne Mountain, there is one type of launch the staff has never seen and doesn't want to see: an intercontinental ballistic missile (ICBM) on its way to the United States. From training sessions, they know exactly what to expect.

The NORAD (North American Aerospace Defense) command center, located deep inside Cheyenne Mountain in Colorado, monitors missile launches around the world. Built in the early 1960s, when the primary nuclear threat was from Soviet long-range bombers, the facility would not withstand a direct hit from a modern missile. "An ICBM of sufficient magnitude and power would make it Cheyenne Valley," says a military spokesman.

First, on the center of their monitors, a grid map would appear showing the part of the world from which the missile was launched. A small red circle represents the launch point of the missile. Next, a computerized voice says "Quick alert! Quick alert!" Then an alarm sounds and a light flashes, signaling everyone in the missile-warning center that a desperate race against the clock has begun. A missile launched from a site halfway around the world can reach the United States in less than 25 minutes.

Within a few seconds, computers plot the missile's trajectory. If it is headed for North America, an officer at NORAD alerts a Pentagon task force and the White House. Major Barry Venable, an Army spokesman for the Cheyenne Mountain operations center, says the President has exactly two options at that point. He can order a massive nuclear counterattack against the aggressor nation. Or he can wait as the incoming missile, or many missiles, vaporizes a city, or many cities, which will happen whether he retaliates or not. Given the current state of technology, no other option exists. The United States cannot shoot down one ICBM, let alone hundreds of them. There are no weapons that can hit a slender projectile traveling at nearly 15,000 miles per hour, about 10 times faster than a bullet. So the obvious question has long been: Can such a defensive weapon be built?

President Bush is betting it can. In a speech at the National Defense University in Washington this past May, the President announced his intention to deploy a system that "could provide limited, but effective, defenses" against ballistic missiles armed with nuclear, biological, or chemical weapons. The defense system, advocates say, is not really intended to thwart an all-out attack by a major nuclear power, which might involve advanced weaponry such as multiple independently targetable reentry vehicles (MIRVs), missiles that bear many warheads. At best they say it could shoot down a few missiles launched by a rogue nation such as North Korea, Iran, or Iraq, as well as missiles accidentally fired from Russia.

The President hopes to make such a missile shield a central accomplishment of his administration. He envisions a complex, multilayered defense that counterattacks with ground-based and sea-based interceptor missiles and aircraft equipped with lasers powerful enough to destroy a rocket in flight. The administration has declared it wants to begin building an interceptor missile base in Alaska that could be operational as early as 2004.

As desirable as even a limited shield might be in an uncertain world, physicists and independent defense analysts who have carefully studied the problem of missile defense argue that the President's plan is deeply flawed, more a product of wishful thinking than sound scientific analysis. "The proponents of missile defense are for the most part totally nontechnical. Or they are defense contractors," says Richard Garwin, a physicist who, with Edward Teller, helped create the world's first hydrogen bomb 50 years ago. "The top levels of the Defense Department are political and managerial, not technical."

And the scientific and technical challenges are far greater than those of any offensive or defensive weapons system ever built. For a missile shield to be effective, every component--radar systems, satellites, missiles, communication networks--would have to perform flawlessly in the heat of combat, says Philip Coyle, who directed the Defense Department's office of testing and evaluation during the Clinton administration. "All the pieces of the system--which are major systems in themselves--have to achieve reliabilities that military equipment rarely ever has."

The flight of an ICBM is a tragic drama that unfolds in three relentless acts: boost phase, midcourse, and reentry.

A three-stage rocket carrying an Exoatmospheric Kill Vehicle was launched on July 14 from Kwajalein Atoll in the western Pacific. So far, missile interceptor tests have not been promising. During one trial, in April 1998, the rocket never left the launchpad due to a programming error. During four subsequent tests since October 1999, the target missile was equipped with a homing beacon. In two of the tests, the interceptor missed the dummy warhead completely.Photograph courtesy of Boeing

The boost phase is by far the best chance to shoot down an enemy ICBM, but it lasts at most 300 seconds, the time a missile takes to clear the atmosphere. A rocket rising into the air presents a clear target, and satellites can easily spot its hot exhaust plume. Two very different sorts of weapons are being considered for this stage of defense, and neither will be ready for at least several years: ship-launched interceptor missiles and an exotic airborne laser.

Of these two weapons, only the airborne laser is in active development. Boeing, Lockheed-Martin, and TRW are trying to equip a modified 747 jumbo jet with up to 17 lasers. In theory, several of these aircraft would fly constant patrol above pariahs like Iraq and North Korea, ready to shoot down a missile within a minute or two of its launch.

Unlike the colored, flashing death rays of science-fiction movies, the airborne laser consists of infrared light, invisible to human eyes. A reaction between rather ordinary chemicals--hydrogen peroxide and potassium hydroxide, the ingredients in hair bleach and Drano--generates the laser. But to produce the megawatt beam that would be needed to bring down a rocket requires linking many such lasers together in a series and sparking them with tons of the chemicals. Plans call for a 747 equipped with 14 laser modules, each weighing about 6,000 pounds when full. The aircraft would also carry three more lasers: one to track the missile, a second to measure atmospheric turbulence, and a third to illuminate the exact target. To compensate for atmospheric effects, a flexible, mechanically controlled mirror that can change its shape every few thousandths of a second will focus and aim the killer beam.

The beam would not immediately explode its target. Instead, after several seconds, heat from the laser would crack the outer surface of one of the three stages of the booster rocket that propels the warhead into space, weakening it. Pressure from the liquid fuel within the stage would then rupture the rocket, causing a catastrophic explosion. No one knows if the idea can be made to work. The first test is not scheduled until 2003, and the project is not expected to be operational for at least seven years.

In the meantime, the sheer bulk of the laser equipment that must be installed on the 747 poses daunting problems. Since the atmosphere would partially absorb and scatter the laser beam, the 747 would have to fly quite high to make sure the beam remained powerful enough to destroy the missile. "A 747 that is heavily loaded doesn't like to fly high," says Philip Coyle. "And you've also got to get close to the forward edge of battle, because your laser power isn't going to be such that it can propagate through many miles of atmosphere. If you're close to the enemy, you make a pretty inviting target yourself."The chemicals that fuel the laser present other problems, says Coyle: "The laser puts out caustic gases in the chemical process. The compatibility of all these materials with an aircraft has yet to be worked out."

If the airborne laser can be built and made to work, it's reasonable to assume an enemy will build defenses to protect its missile from just such an attack. "We're an open society, and we tell people what we're doing," Coyle says. "If the enemy saw that we were doing tests of such a system, and it looked as if it might work, they could foil us pretty easily by just putting a reflective surface of paint on the missile." The reflective surface would need to withstand the laser's heat for just a minute or so, providing enough time for the missile to move into space, well beyond the laser's range.

"One of the dilemmas involved in creating a missile-defense system is that the offense always has the advantage," says Coyle. "They can always figure out how to beat it."

Ted Postol, a nuclear physicist at MIT and a former advisor to the Pentagon on ballistic missile technology, says there is little chance the airborne laser will be ready in 2008 as its developers say. "There are lots and lots of technical questions that are unresolved," he says. "My guess is that it is decades away--if it can be built at all." Still more dubious are plans for space-based lasers mounted on remote-controlled satellites designed to attack ICBMs during the boost phase. Even the most die-hard supporters of that research say space-based lasers won't be defending the United States in the foreseeable future.

The one means of defense that stands the best chance of actually destroying an ICBM in flight--interceptor missiles launched from Navy warships--is still on the drawing board. An interceptor is designed to destroy an ICBM simply by crashing into it. But the deployment of ship-based interceptors is rife with practical problems. For a ship's missiles to have a chance to catch up with and hit an enemy missile, the ship would have to be stationed within a few hundred miles of enemy territory, making it vulnerable to attack. And the interceptors themselves would have to be much faster than any missile in existence, which means they would also be much larger in order to carry the fuel needed to power their five-mile-per-second flight. "It would take whole new rockets that are fatter, longer, and faster than anything the Navy has now," says Coyle. "In addition, this stuff won't fit in the launch tubes on existing Navy ships."

Click on image to enlarge (68k)Illustration by Matt Zang

Like airborne lasers, a ship-launched interceptor would have less than 300 seconds to destroy a missile before it went completely out of range. That leaves little time--perhaps no time--for the President to be consulted about a decision to fire. Coyle says: "You might have to shoot it down within as little as two minutes, certainly within three or four. Otherwise the enemy rocket has cleared the atmosphere. So the decision to fire has got to be computerized, because there just isn't enough time to contemplate what you're doing. The President isn't going to be in the loop; the secretary of defense or the national security advisor wouldn't be in the loop."

The lack of civilian or even human control over such a decision would be unprecedented. The destruction of an enemy warhead could cause debris--including chemical and biological weapons, if aboard--to rain down on neighboring countries. And there is a far greater danger: The interceptor could hit and destroy the enemy rocket without destroying the warhead. In that case the warhead would career off course and land who knows where.

Furthermore, there is little guarantee that an interceptor could hit the target. Coyle emphasizes how hard it is to shoot down combat aircraft that travel at a fraction of an ICBM's speed. "With air defenses--shooting down planes--you're doing really good if you can hit 25 percent of your targets," he says.

When the targets are missiles instead of planes, the statistics are sobering. After the Gulf War against Iraq, Postol analyzed the Army's claims about the Patriot missile's performance against Iraqi missiles launched at Israel. Although the exact number of SCUD- Patriot encounters during the war is classified, there are thought to have been about 80. The Army initially claimed that Patriots hit 96 percent of their targets. Postol found, as did the Government Accounting Office in a separate study, that the Patriots hit at most 6 percent--fewer than five out of 80. In fact, Postol could not confirm that any Patriot missile ever actually hit any SCUD missile.

When an ICBM's booster rocket burns out and falls back to Earth, the warhead travels through the vacuum of space solely under the influence of gravity. This midcourse phase, which lasts about 15 minutes, is the longest of the three phases of an ICBM's flight. To shoot down an ICBM in midcourse, President Bush would field as many as 250 missiles in Alaska and North Dakota, possibly complemented by sea-launched interceptors. Ground-based radar and space-based infrared satellites would guide each interceptor part of the way toward its target. Onboard tracking systems would take over during the final seconds as the interceptor homed in on the enemy warhead. As with the sea-based interceptors proposed for the boost phase of defense, 51-inch-long missiles would destroy the warhead simply by colliding with it.

What if an interceptor misses? Lieutenant General Ronald Kadish, head of the Ballistic Missile Defense Organization, argued in testimony before the Senate Armed Services Committee recently that there should be sufficient time during the midcourse phase to fire additional interceptors at a warhead. "Multiple shots at the target give a higher probability of being able to hit it," he said.

But critics argue that Kadish underestimates enemy countermeasures. During the midcourse phase, the tracking systems guiding an interceptor would probably have to cope not with a single target but with hundreds. The warhead is housed during its ascent in a protective shell called a bus. Once in space, the bus opens, releasing not only the warhead but also numerous decoys to fool satellites and radar networks. Without air to slow them down, all the objects, regardless of size, shape, or weight, travel at about 15,000 miles per hour, clustered around the warhead. Instead of having to destroy an individual missile, as in the boost phase, the interceptor must pick out the real warhead.

Kadish insists that the critics are too pessimistic. "I do not share the assessment that what we are attempting to accomplish with our system is in any way impossible," he told Senate Armed Services Committee members. "Over the coming months and years, I believe program results can speak for themselves in responding to the criticism that the [missile shield] cannot operate as designed against the projected countermeasure threat that a state of concern might pose."

A) A one-third-mile-long tunnel leads to the NORAD missile-warning center inside Cheyenne Mountain, where the main entrance

B) is equipped with two steel blast doors, including one about 25 yards down the hall. Each door weighs 25 tons and is closed by a chain connected to a pneumatic motor.

C) A screen image from a command-center computer console shows the type of data NORAD officials will see if an ICBM is launched from North Korea. The red dot is the launch point and the quadrilateral marks the area potentially threatened by the missile, based on calculations of its range and azimuth.

Coyle, Postol, and other experts do not share Kadish's faith. Both the United States and Russia have designed missiles that contain decoys. There is no reason, says Coyle, why North Korea, Iran, or Iraq could not equip missiles with decoys too. "If they're smart enough to make ICBMs with sophisticated guidance systems and all the rest, I think they can figure out how to make decoys," says Coyle.

The simplest decoys are Mylar balloons, just like those sold in supermarkets, except that each one is about the size of a house. They reflect radar and present tracking stations with hundreds of signals. Besides balloons, the bus might spew out millions of pieces of half-inch-long radar-reflecting wires. Clouds of this chaff, as it is called, would envelop the warhead, making its exact location very difficult to determine. An interceptor could sail through the cloud without touching the warhead. Other chaff clouds would be empty, creating more false targets.

The radar stations and satellites the United States uses can track the launch and flight of a missile very accurately, but they cannot distinguish decoys from targets. To do that, the missile-defense system must rely on a powerful new radar, called X-band, now being tested on Kwajalein Atoll in the western Pacific. Like all radar, it sends out electromagnetic pulses and detects pulses that reflect back off the targets. X-band radar has a much shorter wavelength than any existing radar--just over an inch long--which allows it to resolve details of range, size, and shape that remain invisible to standard radar. The prototype is reported to be able to detect objects as far away as 1,200 miles. When the radar is fully developed, that range is expected to increase to about 2,400 miles. Several X-band radar stations would have to be built around the world.

Coyle points out that the X-band radar's high resolution will create problems of its own. The wavelength is so short, he says that it will reflect off rain or hail in the line of sight between the radar and the objects it tracks. And the radar has been anything but foolproof in early tests (see "Anatomy of a Test," below).

Plans call for X-band radar to work in concert with a fleet of two dozen or so new infrared satellites, which have yet to be built. The program is behind schedule and over budget. In theory, X-band radar would be able to tell a warhead from decoy balloons because the round balloons and cone-shaped warhead would all create unique radar reflections. The satellites would find the target by measuring the amount of infrared radiation--heat--emitted by the warhead and the decoys; a 1,000-pound warhead would remain warmer as it traveled through the cold vacuum of space than would the light balloon decoys, which would rapidly cool. All of this information would be sent to a command center, most likely Cheyenne Mountain, which would automatically relay it to the computers on board the interceptor missile.

But even the most advanced detectors could be foiled by the simplest of countermeasures. Instead of using round balloons, an enemy might make cheap, detailed balloon replicas of the warhead, so the radar reflections from the decoy and the warhead would look identical to the X-band receivers. Small, battery-powered heaters placed inside the decoys could trick infrared satellites into believing they had spotted a real, warm warhead. Alternatively, the warhead itself could be cooled by sheathing it with an insulating aluminum shell filled with liquid nitrogen, hiding it from the infrared satellites (see "How to Hide a Missile," below). An enemy could even enclose the real warhead in a balloon, which would inflate when released from the bus, making the warhead indistinguishable from the decoys. Even if the interceptor managed to hit the house-sized balloon containing the refrigerator-sized warhead, it might just puncture the balloon and sail right past the warhead.

Postol points out another potential obstacle. Should the X-band radar somehow manage to single out the warhead from a herd of decoys, that information might be useless for the interceptor. "If the radar correctly identifies an object as a warhead, it doesn't mean that the kill vehicle will know which object to home in on," he says. The problem is that once the radar has relayed its information to the interceptor, the targets, traveling at about five miles per second, will have moved. "By the time the kill vehicle encounters the targets, they will have been remixed up," says Postol.

Closing on the flock of balloons, chaff, and warhead at more than 15,000 miles per hour, the interceptor would have to rely on its own sensors and would likely miss the warhead. Moreover, an enemy would not have to arm its missiles with nuclear warheads and decoys to defeat a defense shield. Instead, it could opt to load one warhead with a hundred or more small "bomblets," each packed with a biological weapon like anthrax spores. Dropped on a city and dispersed by the wind, they might be more devastating than any nuclear weapon. One hundred bomblets could release 440 pounds of anthrax over a city, enough to kill more than 100,000 people. A nuclear weapon built with third-world technology might kill 60,000. A missile-defense shield would be utterly helpless against such a threat--the bomblets would be far too small and numerous for any interceptors to destroy or even see.

A brief but climactic third act--reentry--follows the midcourse phase of the ICBM's flight. This final phase lasts only 40 seconds, and the chance of a successful interception at this point is nil. Although the light decoys will have burned up in the atmosphere, leaving the enemy warhead alone at last, the interceptors probably would not see it. Below a height of 80 miles, the heat of friction with the atmosphere blinds heat-seeking sensors. In any case, interceptors might still face hundreds of germ-bearing bomblets, not a lone warhead.

The 15 buildings inside the Cheyenne Mountain complex are supported by 1,319 springs, each weighing 1,000 pounds. They were designed to cushion the effects of a nuclear explosion. "One great irony is that the yield and accuracy of nuclear weapons systems improved quickly enough to render the facility potentially vulnerable the day it was completed in 1965," says NORAD spokesman Major Barry Venable.

In describing the technical challenge of destroying an ICBM hurtling through space, some scientists have likened the task to hitting a bullet with a bullet. "We know how to hit a bullet with a bullet," Coyle says. "So that's not really the analogy. It's like when you were a kid and somebody threw a ball at you. You could stop that. But if they threw a handful of rocks you couldn't stop them all. That's the problem. We won't have enough bullets."

When the Cheyenne Mountain complex was completed in 1965, it was designed to withstand the nearby explosion of a 30-megaton warhead. Today's weapons are many times more powerful, and the operations center is no longer impregnable. These days, as the staff at Cheyenne Mountain continues to watch for the unthinkable, just as they have since the height of the cold war, they don't bother to close the blast doors anymore. If the unthinkable missile is spotted, the doors can be shut within 30 seconds, locking 800 people inside with enough food, water, heat, electricity, and filtered air to enable them to survive for about 30 days.

"During the cold war we used to keep them closed all the time--except during shift changes," says Venable. Now he and his colleagues hope they will never be closed again.

What is the threat?

By far the greatest threat of an attack is still posed by Russia, which possesses about 5,200 warheads mounted on missiles, probably 2,000 fewer than the United States. An accidental launch--or an unauthorized one by a rebellious officer--would probably involve at least a few dozen warheads, and possibly thousands. The commander of a single Russian submarine, for example, could fire as many as 64 warheads at the United States, enough to kill tens of millions of people. Such an attack would completely overwhelm any missile-defense system being contemplated.

China is believed to have no more than 20 ICBMs that could reach the United States. For now, because China apparently stores its warheads and rocket fuel separately from its missiles, there is little chance of an accidental launch. China also has one submarine armed with nuclear weapons, but it usually remains close to the Chinese mainland, and its missiles don't have the range to reach American territory. But the construction of an American missile-defense system could make China adopt a more aggressive policy and place fully armed missiles on high-alert status.

Among third-world countries considered hostile to the United States, only North Korea has actually launched a multistage rocket--the Taepo Dong I, which flew over Japan in 1998 before crashing into the Pacific. Unclassified U.S. government intelligence reports speculate that within 10 years North Korea, Iran, and Iraq might be able to build missiles that could hit the United States, but no one knows whether those countries will actually pursue such a program. ICBMs are expensive to develop, so a hostile third-world country might choose instead to smuggle biological weapons into the United States and then release them. Or a rogue state could stow a nuclear weapon on a cargo ship and detonate the weapon in an American harbor. Any of these strategies could be anonymous as well as potentially more destructive than an ICBM. — T. F.

How to hide a missile

To find and destroy an ICBM traveling 10 times faster than a bullet 140 miles above Earth's surface requires exquisite timing. With an interceptor and an ICBM approaching each other at closing speeds measured in miles per second, a half-second delay in spotting the target could make the interceptor miss by thousands of feet. For most of its flight, the interceptor will be guided toward the ICBM by radar stations on Earth and by satellites. Although the exact capability of the interceptor's infrared instruments is classified, in flight tests the interceptor apparently detected its target at a range of about 450 miles. Those targets, however, were warm and stood out against the cold background of space. But an enemy could enclose a warhead in a thin metallic shell filled with liquid nitrogen; physicist Richard Garwin says such technology is not challenging. The liquid nitrogen would cool the warhead to 77 Kelvin, or -321 degrees Fahrenheit. At that temperature, the interceptor's sensors wouldn't see the target until it was only about half a mile away. Could the interceptor use its small maneuvering rockets to correct its course at such a short distance and hit the ICBM? As the diagram above shows, assuming the interceptor needed to change course by just 60 feet, it would need to accelerate at a rate of more than 400 g's, far beyond its ability. — T. F.

Anatomy of a test

At 10:40 p.m. eastern time on July 14, a Minuteman II ICBM carrying a dummy warhead took off from Vandenberg Air Force Base near Los Angeles. Exactly 21 minutes later an interceptor missile roared into the sky from Kwajalein Atoll in the western Pacific, some 4,800 miles away. When a flash of light indicating that the missiles had collided appeared on closed-circuit televisions in the Pentagon at 11:09 p.m., supporters of the missile-defense program were jubilant. What went largely unreported was that the dummy warhead had a target beacon on board. "That's not a bad thing to do for a first test, but it was not a demonstration of something you could deploy," says Philip Coyle, who headed the Pentagon's department of testing and evaluation during the Clinton administration. "Presumably a country that attacks us wouldn't put beacons on their missiles."

Photograph courtesy of Boeing

More problematic was the performance of the new X-band radar station on Kwajalein, which was supposed to track the last leg of the target's flight, up to and including the collision with the interceptor. Debris completely confused the radar. "Every time there's a stage separation, there are belts and straps and different objects that come loose," says Coyle. Reflections from that debris prompted the X-band radar to indicate that the interceptor had missed the target. This raises questions about how well X-band radar would deal with simple enemy countermeasures, such as dumping millions of pieces of radar-reflecting wire, or chaff, around the warhead. "A chaff cloud would be a great countermeasure," says Coyle. "Lord knows the North Koreans are smart enough to do that." — T. F.

Targeting Rogue Missiles

If Iraq, Iran, or North Korea launched an ICBM toward the United States, the best time for an in-flight interception would be during the 300-second-long boost phase. And the best way to do that, says physicist Richard Garwin, would be for the United States to have ground-based interceptor missiles in Turkey ready to shoot down Iraqi ICBMs, have interceptors located near the Caspian Sea to destroy Iranian missiles, and place missiles in Russia to foil any launches from North Korea. That would of course entail a joint effort with Russia. But one clear advantage of the plan is that Russia is not likely to feel threatened by interceptors clearly targeted at rogue nations. Moreover, ground-based interceptors in Russia and Turkey would not involve exotic, unproven technologies like the airborne laser that is a key element of the Bush administration's proposed missile shield. The interceptors now being developed for the national missile-defense system are designed to have a top speed of 5.3 miles per second--compared with an ICBM's speed of 4.7 miles per second--so boost-phase interceptors would have to be modified to reach their top speed within 100 seconds of an ICBM's launch. Such a system is far more likely to be successful than an exotic infrared laser system. With such a system in place, the United States might have a chance of defending itself and other countries against a rogue-nation missile attack. And, as Garwin wrote in the Bulletin of the Atomic Scientists, "it would be far less costly than the proposed national missile-defense system."— T. F.

Photograph courtesy of Boeing

The Center for Defense Information has an informative site on some of the basics behind national missile defense:

The Union of Concerned Scientists reports on a number of possible countermeasures to a national missile defense, as well as the viability of the proposed program:

NORAD's Web site:

1 free article left
Want More? Get unlimited access for as low as $1.99/month

Already a subscriber?

Register or Log In

1 free articleSubscribe
Discover Magazine Logo
Want more?

Keep reading for as low as $1.99!


Already a subscriber?

Register or Log In

More From Discover
Recommendations From Our Store
Shop Now
Stay Curious
Our List

Sign up for our weekly science updates.

To The Magazine

Save up to 40% off the cover price when you subscribe to Discover magazine.

Copyright © 2024 Kalmbach Media Co.